IMX6 DDR3 Rules PDF

Hardware Development Guide
for i.MX 6Quad, 6Dual, 6DualLite,

6Solo Families of
Applications Processors
IMX6DQ6SDLHDG
Rev 1
06/2013
How to Reach Us: Information in this document is provided solely to enable system and software
implementers to use Freescale products. There are no express or implied copyright
Home Page:
freescale.com licenses granted hereunder to design or fabricate any integrated circuits based on the
Web Support: information in this document.

freescale.com/support Freescale reserves the right to make changes without further notice to any products
herein. Freescale makes no warranty, representation, or guarantee regarding the
suitability of its products for any particular purpose, nor does Freescale assume any
liability arising out of the application or use of any product or circuit, and specifically
disclaims any and all liability, including without limitation consequential or incidental
damages. “Typical” parameters that may be provided in Freescale data sheets and/or
specifications can and do vary in different applications, and actual performance may
vary over time. All operating parameters, including “typicals,” must be validated for each
customer application by customer’s technical experts. Freescale does not convey any
license under its patent rights nor the rights of others. Freescale sells products pursuant
to standard terms and conditions of sale, which can be found at the following address:
http://www.reg.net/v2/webservices/Freescale/Docs/TermsandConditions.htm.
Freescale and the Freescale logo are trademarks of Freescale Semiconductor, Inc.,
Reg. U.S. Pat. & Tm. Off. All other product or service names are the property of their
respective owners. ARM and Cortex are registered trademarks of ARM Limited.
© 2013 Freescale Semiconductor, Inc.
Document Number: IMX6DQ6SDLHDG

Rev 1, 06/2013
Contents
Paragraph Page
Number Title Number
Chapter 1
About This Book
1.1 Overview.......................................................................................................................... 1-1

1.2 Devices supported............................................................................................................ 1-1
1.3 Essential reference ........................................................................................................... 1-1
1.4 Suggested reading ............................................................................................................ 1-1
1.4.1 General Information..................................................................................................... 1-1
1.5 Related documentation .................................................................................................... 1-2
1.6 Conventions ..................................................................................................................... 1-2
1.7 Signal conventions ........................................................................................................... 1-3
1.8 Acronyms and abbreviations ........................................................................................... 1-3
Chapter 2
Design Checklist
2.1 Design checklist overview ............................................................................................... 2-1

2.2 Design checklist tables..................................................................................................... 2-1
2.3 Bus isolation circuit ....................................................................................................... 2-11
2.4 DDR reference circuit .................................................................................................... 2-11
2.5 I2C address..................................................................................................................... 2-12
2.5.1 I2C clock speed and division factors (IFDR)............................................................. 2-13
2.6 JTAG signal termination ................................................................................................ 2-16
2.7 Oscillator tolerance ........................................................................................................ 2-16
2.8 Unused analog interfaces ............................................................................................... 2-16
Chapter 3
i.MX 6 Series Layout Recommendations
3.1 Basic design recommendations........................................................................................ 3-1

3.1.1 Fanout illustrations ...................................................................................................... 3-3
3.1.2 Placing decoupling capacitors ..................................................................................... 3-4
3.2 Stackup recommendations ............................................................................................... 3-4
3.3 DDR connection information........................................................................................... 3-6
3.4 DDR routing rules............................................................................................................ 3-8
3.5 Routing considerations .................................................................................................... 3-9
Hardware Development Guide for i.MX 6Dual/6Quad and i.MX 6Solo/6DualLite Applications Processors, Rev. 1
Freescale Semiconductor iii
Contents
Paragraph Page
Number Title Number
3.5.1 Swapping data lines ................................................................................................... 3-10
3.5.2 DDR3 (64 bits) T topology considerations................................................................ 3-10
3.5.3 DDR3 (64 bits) Fly-by topology considerations........................................................ 3-10
3.5.4 2-Gigabyte recommendations .................................................................................... 3-10
3.5.5 4-Gigabyte recommendations .................................................................................... 3-11
3.5.6 Four chips T topology routing examples ................................................................... 3-13
3.5.7 Eight chips fly-by topology routing examples........................................................... 3-17
3.5.8 High speed signal routing recommendations............................................................. 3-25
3.5.9 Ground plane recommendations ............................................................................... 3-25
3.6 DDR power recommendations....................................................................................... 3-28
3.7 PCI Express interface recommendations ...................................................................... 3-29
3.7.1 PCI Express general routing guidelines..................................................................... 3-29
3.7.2 PCI Express coupling lane......................................................................................... 3-30
3.7.3 Additional resources for PCI Express signal routing recommendations ................... 3-30
3.8 HDMI recommendations ............................................................................................... 3-30
3.9 SATA recommendations ................................................................................................ 3-30
3.10 LVDS recommendations ................................................................................................ 3-30
3.11 USB recommendations .................................................................................................. 3-31
3.12 Impedance signal recommendations .............................................................................. 3-31
3.13 Reference resistors ......................................................................................................... 3-32
3.14 ESD and radiated emissions recommendations ............................................................. 3-33
3.15 Component placement recommendations ...................................................................... 3-33
3.16 Reducing skew and phase problems in deferential pairs traces..................................... 3-33
3.17 Guideline for power net electrical performance ............................................................ 3-35
Chapter 4
Requirements for Power Management
4.1 Power management requirements overview .................................................................... 4-1

4.1.1 Voltage domains overview........................................................................................... 4-1
4.1.2 PF0100 overview ......................................................................................................... 4-1
4.2 Requirements for a generic interface between chip and PF0100..................................... 4-1
4.3 i.MX6 internal regulators................................................................................................. 4-4
4.4 Connection diagrams ....................................................................................................... 4-6
4.5 Video power recommendations........................................................................................ 4-8
Chapter 5
Using the Clock Connectivity Table
5.1 Root clocks ...................................................................................................................... 5-1

5.2 Waking the core up from stop mode ................................................................................ 5-2
iv Freescale Semiconductor
Contents
Paragraph Page
Number Title Number
Chapter 6
Using the IOMUX Design Aid
6.1 Compatibility across the i.MX 6 series families of processors........................................ 6-2

6.2 Application requirements................................................................................................. 6-2
6.3 IOMUX tool version ........................................................................................................ 6-2
6.4 IOMUX tool location....................................................................................................... 6-2
6.5 Learning to use the IOMUX tool ..................................................................................... 6-2
Chapter 7
Configuring JTAG Tools
7.1 JTAG tool requirements ................................................................................................... 7-1

7.2 Extra JTAG functionality ................................................................................................. 7-1
7.3 Updating your RealView ICE .......................................................................................... 7-2
7.4 Defining the JTAG chain ................................................................................................. 7-3
7.5 Reading a register with RealView Debugger v4.1........................................................... 7-6
7.6 CoreSight Base address references .................................................................................. 7-8
Chapter 8
Avoiding Board Bring-up Problems
8.1 Using a current monitor to avoid power pitfalls .............................................................. 8-1

8.2 Using a voltage report to avoid power pitfalls................................................................. 8-1
8.3 Checking for clock pitfalls............................................................................................... 8-2
8.4 Avoiding reset pitfalls ...................................................................................................... 8-3
8.5 Sample board bring-up checklist ..................................................................................... 8-4
Chapter 9
Understanding the IBIS Model
9.1 IBIS structure and content ............................................................................................... 9-1

9.2 Header Information.......................................................................................................... 9-2
9.3 Component and pin information ...................................................................................... 9-2
9.4 Model information ........................................................................................................... 9-4
9.4.1 IV information ............................................................................................................ 9-5
9.4.2 VT information ............................................................................................................ 9-5
9.4.3 Golden Model VT information .................................................................................... 9-7
9.5 Freescale naming conventions for model names and usage in i.MX6 IBIS file.............. 9-8
9.5.1 [Model Selector] ddr.................................................................................................... 9-8
9.5.1.1 DDR [Model Selector]............................................................................................. 9-8
Freescale Semiconductor v
Contents
Paragraph Page
Number Title Number
9.5.1.2 RGMII...................................................................................................................... 9-9
9.5.2 [Model Selector] gpio ................................................................................................ 9-10
9.5.3 [Model Selector] lvds................................................................................................. 9-11
9.5.4 [Model Selector] mlb ................................................................................................. 9-11
9.5.5 [Model Selector] USB ............................................................................................... 9-11
9.5.6 List of pins not modeled in the i.MX6 IBIS file........................................................ 9-12
9.6 Quality assurance for the IBIS models .......................................................................... 9-12
9.7 IBIS usage...................................................................................................................... 9-13
9.8 References...................................................................................................................... 9-13
Chapter 10
Using the Manufacturing Tool
10.1 Overview........................................................................................................................ 10-1

10.2 Feature summary............................................................................................................ 10-1
10.3 Other references ............................................................................................................. 10-2
Chapter 11
Using BSDL for Board-level Testing
11.1 BSDL overview ............................................................................................................. 11-1

11.2 How BSDL functions..................................................................................................... 11-1
11.3 Downloading the BSDL file .......................................................................................... 11-1
11.4 Pin coverage of BSDL ................................................................................................... 11-1
11.5 Boundary scan operation ............................................................................................... 11-2
11.6 I/O pin power considerations ......................................................................................... 11-2
Chapter 12
Using the RMII Interface
12.1 Overview........................................................................................................................ 12-1

12.2 Configuring the RMII signal connections ..................................................................... 12-2
12.3 Generating the reference clock ...................................................................................... 12-4
12.4 Generating the reference clock on chip ......................................................................... 12-4
12.4.1 Using the GPIO_16 pin to generate the reference clock ........................................... 12-5
12.4.2 Using RGMII_TX_CTL to generate the reference clock .......................................... 12-6
12.5 Using an external clock.................................................................................................. 12-7
Appendix A
Development Platforms
vi Freescale Semiconductor
Contents
Paragraph Page
Number Title Number
Appendix B
Revision History
Freescale Semiconductor vii
Contents
Paragraph Page
Number Title Number
viii Freescale Semiconductor
Chapter 1
About This Book
1.1 Overview
This document’s purpose is to help hardware engineers design and test their i.MX 6 series processor based
designs. It provides information on board layout recommendations, design checklists to ensure first-pass
success and ways to avoid board bring-up problems. It also provides information on board-level testing
and simulation such as properly configuring JTAG tools, using BSDL for board-level testing, using the
IBIS model for electrical integrity simulation and more.
Engineers are expected to have a working understanding of board layouts and terminology, IBIS modeling,
BSDL testing and common board hardware terminology.
This guide is released along with relevant device-specific hardware documentation such as datasheets,
reference manuals and application notes available on www.freescale.com.
1.2 Devices supported

This Hardware Developer’s Guide currently supports the i.MX 6Quad, 6Dual, 6DualLite and 6Solo
families of application processors.
1.3 Essential reference

This guide is intended as a companion to the i.MX 6 series chip reference manuals and data sheets. For
reflow profile and thermal limits during soldering, see application note AN3298. These documents are
available on www.freescale.com.
1.4 Suggested reading

This section lists additional reading that provides background for the information in this manual as well as
general information about the architecture.
1.4.1 General Information

The following documentation provides useful information about the ARM processor architecture and
computer architecture in general:
For information about the ARM Cortex-A9 processor see:
http://www.arm.com/products/processors/cortex-a/cortex-a9.php
• Computer Architecture: A Quantitative Approach (Fourth Edition) - by John L. Hennessy and
David A. Patterson
Freescale Semiconductor 1-1
About This Book
• Computer Organization and Design: The Hardware/Software Interface (Second Edition), by

David A. Patterson and John L. Hennessy
The following documentation provides useful information about high-speed board design:
• Right the First Time- A Practical Handbook on High Speed PCB and System Design -
Volumes I & II - Lee W. Ritchey (Speeding Edge) - ISBN 0-9741936- 0-72
• Signal and Power Integrity Simplified (2nd Edition) - Eric Bogatin (Prentice Hall)- ISBN
0-13-703502-0
• High Speed Digital Design- A Handbook of Black Magic - Howard W. Johnson & Martin
Graham (Prentice Hall) - ISBN 0-13-395724-1
• High Speed Signal Propagation- Advanced Black Magic - Howard W. Johnson & Martin
Graham - (Prentice Hall) - ISBN 0-13-084408-X
• High Speed Digital System Design- A handbook of Interconnect Theory and Practice - Hall,
Hall and McCall (Wiley Interscience 2000) - ISBN 0-36090-2
• Signal Integrity Issues and Printed Circuit Design - Doug Brooks (Prentice Hall) ISBN
0-13-141884-X
• PCB Design for Real-World EMI Control - Bruce R. Archambeault (Kluwer Academic
Publishers Group) - ISBN 1-4020-7130-2
• Digital Design for Interference Specifications- A Practical Handbook for EMI Suppression -
David L. Terrell & R. Kenneth Keenan (Newnes Publishing) - ISBN 0-7506-7282-X
• Electromagnetic Compatibility Engineering- Henry Ott (1st Edition - John Wiley and Sons) -
ISBN 0-471-85068-3
• Introduction to Electromagnetic Compatibility - Clayton R. Paul (John Wiley and Sons) - ISBN
978-0-470-18930-6
• Grounding & Shielding Techniques - Ralph Morrison (5th Edition - John Wiley & Sons) - ISBN
0-471-24518-6
• EMC for Product Engineers - Tim Williams (Newnes Publishing) - ISBN 0-7506- 2466-3
1.5 Related documentation

Freescale documentation is available from the sources listed on the back page of this guide.
Additional literature is published as new Freescale products become available. For a current list of
documentation, see www.freescale.com.
1.6 Conventions
This document uses the following notational conventions:
Courier Used to indicate commands, command parameters, code examples, and file and
directory names.
Italics Italics indicates command or function parameters
Bold Function names are written in bold.
1-2 Freescale Semiconductor
About This Book
cleared/set When a bit takes the value zero, it is said to be cleared; when it takes a value of
one, it is said to be set.
mnemonics Instruction mnemonics are shown in lowercase bold
Book titles in text are set in italics
sig_name Internal signals are written in all lowercase
nnnn nnnnh Denotes hexadecimal number
0b Denotes binary number
rA, rB Instruction syntax used to identify a source GPR
rD Instruction syntax used to identify a destination GPR
REG[FIELD] Abbreviations for registers are shown in uppercase text. Specific bits, fields, or
ranges appear in brackets. For example, MSR[LE] refers to the little-endian mode
enable bit in the machine state register.
x In some contexts, such as signal encodings, an unitalicized x indicates a don’t
care.
x An italicized x indicates an alphanumeric variable
n, m An italicized n indicates a numeric variable
NOTE
In this guide, notation for all logical, bit-wise, arithmetic, comparison, and
assignment operations follow C Language conventions.
1.7 Signal conventions

PWR_ON_RESET An overbar indicates that a signal is active when low
_b, _B Alternate notation indicating an active-low signal
signal_name Lowercase italics is used to indicate internal signals
1.8 Acronyms and abbreviations

The following table defines the acronyms and abbreviations used in this document.
Table 1: Definitions and acronyms

Term Definition
ARM® Advanced RISC machines processor architecture
BGA Ball grid array package
BOM Bill of materials
BSDL Boundary scan description language
CAN Flexible Controller Area Network peripheral
CCM Clock Controller Module
About This Book
Table 1: Definitions and acronyms (continued)

CSI MIPI camera serial interface
DDC VESA Data Display Channel
DDR Dual data rate DRAM
DDR3 DDR3 DRAM
DDR3L Low voltage DDR3 DRAM
DDR3U Ultra low voltage DDR3 DRAM
DRAM Dynamic random access memory
DSI MIPI display serial interface
ECSPI Enhanced Configurable SPI peripheral
EDID Extended Display Identification Data
EIM External Interface Module
ENET 10/100/1000-Mbps Ethernet MAC peripheral
EPIT Enhanced Periodic Interrupt Timer peripheral
ESR Equivalent series resistance (of a crystal)
FSL Freescale Semiconductor
GND Ground
GPC General Power Controller
GPIO General-purpose input/output
HDCP High-bandwidth Digital Content Protection
HDMI High-definition multimedia interface
I2C Inter-integrated circuit interface
IBIS Input output buffer information specification
IOMUX i.MX6 chip-level I/O multiplexing
JTAG Joint Test Action Group

KPP Keypad Port peripheral
LDB LVDS Display bridge
LDO Low drop-out regulator
LPCG Low power clock gating
LPDDR2 Low-power DDR2 DRAM
LVDS Low-voltage differential signaling
MLB MediaLB 150 peripheral
MMDC Multi Mode DDR Controller
ODT On-die termination
OTP One-time programmable
About This Book
Table 1: Definitions and acronyms (continued)

PCB Printed circuit board
PCIe PCI Express
PCISig Peripheral Component Interconnect Special Interest Group
PMIC Power management integrated circuit
PoP Package-on-package
POR Power-on reset
RAM Random access memory
RGMII Reduced Gigabit Media Independent Interface (Ethernet)
RMII Reduced Media Independent Interface (Ethernet)
ROM Read-only memory
SATA Serial ATA
SDMA Smart Direct Memory Access Controller
UART Universal asynchronous receiver/transmitter
USB Universal Serial Bus
USB OTG USB On-the-go
USB2.0 USB version 2.0 peripheral
VPU Video processing units
About This Book
Chapter 2
Design Checklist
2.1 Design checklist overview
This chapter provides a design checklist for the following i.MX 6 series families of processors:
• i.MX 6Quad
• i.MX 6Dual
• i.MX 6DualLite
• i.MX 6Solo
The design checklist tables (Table 2-1–Table 2-14) contain recommendations for optimal design. Where
appropriate, the checklist tables also provide an explanation of the recommendation so that users have a
greater understanding of why certain techniques are recommended. All supplemental tables referenced by
the checklist appear in sections following the design checklist tables.
See also the application note Common Hardware Design for i.MX 6Dual/6Quad and i.MX
6Solo/6DualLite (AN4397).
2.2 Design checklist tables

Table 2-1. DDR recommendations
Checkbox Recommendation Explanation/supplemental recommendation
1. Connect ZQPAD to an external 240 Ω 1% resistor This is a reference used during DRAM output buffer
to GND. driver calibration.
2. Connect DRAM_VREF to a source that is 50% of • The user may tie DDR_VREF to a precision external
the voltage value of NVCC_DRAM. resistor divider. Shunt each resistor with a
closely-mounted 0.1 μF capacitor. See Table 2-15 for
resistor values. Using resistors with recommended
tolerances ensures the ±2% DDR_VREF tolerance
per the DDR3 specification.
• The user can use a PMIC’s tracking regulator as
used on Freescale reference designs. A tracking
regulator is recommended as a reference for memory
configurations of more than four devices.
Design Checklist
Table 2-1. DDR recommendations (continued)
3. Connect DRAM_RESET to a 10 kΩ 5% pulldown • DDR3: DRAM_RESET should be pulled down to

resistor to GND. meet the JEDEC sequence until the controller is
configured and starts driving. DRAM_RESET should
be kept high when DDR3 enters self-refresh mode.
• LPDDR2: DRAM_RESET should be left
unconnected.
Some Freescale reference designs use a 1% resistor
simply to consolidate the BOM.
DRAM_RESET is an active-low signal.
4. DRAM_SDCKE0 and DRAM_SDCKE1 no longer Both DRAM_SDCKE0 and DRAM_SDCKE1 have

require external 10 kΩ resistors to GND to minimize on-chip pull-downs.
current drain during deep sleep mode (DSM).
5. Make sure that the correct LPDDR2 function is MMDC IO names are for the DDR3 default. When
connected to the correct I/O. Note that this does not LPDDR2 is selected, the I/O name (DDR3 MMDC PAD)
necessarily correspond to the I/O name. does not match with the LPDDR2 functionality. See the
“LPDDR2 and DDR3 pin mux mapping” table in the
“Multi Mode DDR Controller (MMDC)” chapter in your
chip reference manual.
Table 2-2. EIM recommendations for developer’s boot modes
1. When EIM boot signals are used as the system’s Because only resistors are used, EIM bus loads can
EIM signals, other functions, or GPIO outputs after cause current drain, leading to higher (false) supply
boot, use a passive resistor network to select the current measurements. Each EIM boot signal should
desired boot mode for development boards. connect to a series resistor to isolate the bus from the
resistors and/or switchers; see Figure 2-1. Each
configured EIM boot signal sees either a 14.7 kΩ
pulldown or a 4.7 kΩ pullup. For each switch-enabled
pulled-up signal, the supply is presented with a 10 kΩ
current load.
Design Checklist
Table 2-2. EIM recommendations for developer’s boot modes (continued)
2. To reduce incorrect boot-up mode selections, do Using EIM boot interface lines as inputs may result in a
one of the following: wrong boot up due to the source overcoming the pull
• Use EIM boot interface lines only as processor resistor value. A peripheral device may require the EIM
outputs. Ensure EIM boot interface lines are not signal to have an external or on-chip resistor to minimize
loaded down such that the level is interpreted as signal floating.
low during power-up, when the intent is to be a If the usage of the EIM boot signal affects the peripheral
high level, or vice versa. device, then an analog switch, open collector buffer, or
• If an EIM boot signal must be configured as an equivalent should isolate the path. A pullup or pulldown
input, isolate the EIM signal from the target driving resistor at the peripheral device may be required to
source with one analog switch and apply the logic maintain the desired logic level. Review the switch or
value with a second analog switch. Alternately, device data sheet for operating specifications.
peripheral devices with three-state outputs may be
used; ensure the output is high-impedance during
the boot up interval.
3. The BOOT_CFG signals are required for proper See the “System Boot” chapter in your chip reference
functionality and operation and should not be left manual for the correct boot configuration. Note that an
floating. incorrect setting may result from an improper booting
sequence.
Table 2-3. Boot mode input recommendations
1. For BOOT_MODE1 and BOOT_MODE0, use one of Boot inputs BOOT_MODE1 and BOOT_MODE0 each
the following options to achieve logic 0: have on-chip pulldown devices with a nominal value of 100
• Tie to GND through any size external resistor kΩ, a projected minimum of 60 kΩ, and a projected
• Tie directly to GND maximum of 140 kΩ.
• Float Be aware that when these are logic high, current is drawn
For logic 1, use one of the following: from the VDD_SNVS supply.
• Tie directly to the VDD_SNVS_IN rail In production, when on-chip fuses determine the boot
• Tie to the VDD_SNVS_IN rail through an external configuration, both boot mode inputs can be no connects.
resistor 10 kΩ. A value of 4.7 kΩ is preferred in
high-noise environments.
If switch control is desired, no external pulldown
resistors are necessary. Simply connect SPST switches
directly to the VDD_SNVS_IN rail. If desired, a 4.7 kΩ
to 10 kΩ series resistor can be used when current drain
is critical.
Design Checklist
Table 2-4. I2C recommendations
1. Verify the target I2C interface clock rates. The bus can only operate as fast as the slowest peripheral
on the bus. If faster operation is required, move the slow
devices to another I2C port.
2. Verify that the target I2C address range is supported These chips support up to:
and does no conflict with other peripherals. If there is • Three I2C ports for the i.MX 6Quad and 6Dual families
an unavoidable address conflict, move the offending • Four I2C ports for the i.MX 6DualLite and 6Solo
2
device to another I C port. See Table 2-16. families.
If it is undesirable to move a conflicting device to another
I2C port, review the peripheral operation to see if it
supports remapping the address.
3. Do not place more than one set of pullup resistors on This can result in excessive loading. Good design
the I2C lines. practice is to place one pair of pullups only.
Table 2-5. JTAG recommendations
1. Do not use external pullup or pulldown resistors on JTAG_TDO is configured with an on-chip keeper circuit
JTAG_TDO. such that the floating condition is actively eliminated if an
external pull resistor is not present. An external pull
resistor on JTAG_TDO is detrimental. See Table 2-19 for
a summary of the JTAG interface.
2. Ensure that the on-chip pullup/pulldown External resistors can be used with all JTAG signals
configuration is followed. If external resistors are used except JTAG_TDO, but they are not required. See
with JTAG signals, with the exception of JTAG_TDO. Table 2-19 for a summary of the JTAG interface.
For example, do not use an external pulldown on an
input that has an on-chip pullup.
3. JTAG_MOD may be referred to as SJC_MOD in When JTAG_MOD is low, the JTAG interface is configured
some documents. Both names refer to the same signal. for common software debug, adding all the system taps to
JTAG_MOD should be externally connected to GND for the chain.
normal operation in a system. Termination to GND When JTAG_MOD is high, the JTAG interface is
through an external pulldown resistor is allowed. configured to a mode compliant with the IEEE 1149.1
Use ≤ 4.7 kΩ. standard.
Design Checklist
Table 2-6. Power and decouple recommendations
1. Comply with the power-up sequence guidelines as Any deviation from these sequences may result in the
described in the data sheet to guarantee reliable following situations:
operation of the device. • Excessive current during power-up phase
• Prevention of the device from booting
• Irreversible damage to the processor (worst-case
scenario)
2. Do not overload coin cell backup power rail Freescale PMIC PMPF0100 VSNVS regulator is rated to
VDD_SNVS_IN. Note that the following I/Os are supply 400 μA output current under worst-case operating
associated with VDD_SNVS_IN; most inputs have conditions. The VDD_SNVS_IN regulator can supply
on-chip pull resistors and do not require external larger current in transient situations without damaging the
resistors: regulator.
• POR_B – on-chip pullup; see Table 2-8 #1 Concerning i.MX6:
• ONOFF – on-chip pullup; see Table 2-8 #2 • When VDD_SNVS_IN = VDD_HIGH_IN, SNVS
• BOOT_MODE0 – on-chip pulldown; see Table 2-3 domain current is drawn from both equally.
#1 • When VDD_HIGH_IN > VDD_SNVS_IN,
• BOOT_MODE1 – on-chip pulldown; see Table 2-3 VDD_HIGH_IN supplies all SNVS domain current and
#1 current flows into VDD_SNVS_IN to charge a coin cell
• TAMPER – on-chip pulldown battery.
• PMIC_STBY_REQ – push-pull output • When VDD_SNVS_IN > VDD_HIGH_IN,
• PMIC_ON_REQ – push-pull output VDD_SNVS_IN supplies current to SNVS, and some
• TEST_MODE – on-chip pulldown; see Table 2-14 #1 current flows into VDD_HIGH_IN.
Note:VDD_HIGH_IN must be valid (above the internal
detector threshold, 2.4 V typ) for the current flow to
occur. Thus, current flow only happens when
VDD_HIGH_IN is powered to a level below
VDD_SNVS_IN. If VDD_HIGH_IN is off or low, no extra
current is drawn from VDD_SNVS_IN. The whole
circuit assumes it is charging a coin cell and starts
charging when VDD_HIGH_IN is valid. If you are
driving VDD_SNVS_IN with a non-battery power
source, it must be at the same level as VDD_HIGH_IN
or current will flow between them.
• When VDD_SNVS_IN is not powered by a battery, it is
recommended that VDD_SNVS_IN = VDD_HIGH_IN.
If VDD_SNVS_IN is tied to a battery, the battery eventually
discharges to a value equal to that of VDD_HIGH_IN and
never subsequently charges above VDD_HIGH_IN.
The battery chemistry may add restrictions to
VDD_HIGH_IN’s voltage range. External charging
components should be based on the battery
manufacturer's specifications.
Design Checklist
Table 2-6. Power and decouple recommendations (continued)
3. Only one 22 μF bulk capacitor should be connected If the nominal capacitance value is larger than
to each of these on-chip LDO regulator outputs: recommended, power-up ramp time is excessive and
• VDD_ARM_CAP operation cannot be guaranteed. Note that the ramp up
• VDDARM23_CAP time is constant. Larger capacitors mean more inrush
• VDD_SOC_CAP current. Select small capacitors with low ESR (equivalent
• VDD_PU_CAP series resistance).
A 22 μF bulk capacitor must be placed as near as The 22 μF bulk capacitors should be placed as close as
possible with pins/vias. The distance should be less possible to the associated VDD_xx_CAP ball, with trace
than 50mil between bulk cap and VDD_xx_CAP pins. widths and via sizes appropriate to the expected current
Decoupling capacitors such as 0.1 μF or 0.22 μF should draw. A trace length of less than 50 mil is recommended.
also be used. Do not connect any loads to these LDO outputs:
VDDARM_CAP, VDDARM23_CAP, or VDDPU_CAP.
VDDSOC_CAP is restricted to MX6 loads.
4. Only one 10 μF bulk capacitor should be connected If the nominal capacitance value is larger than
to each of these on-chip LDO regulator outputs: recommended, power-up ramp time is excessive and
• VDD_HIGH_CAP operation cannot be guaranteed. Select small capacitors
• NVCC_PLL_OUT with low ESR.
• VDD_USB_CAP These LDOs should only be used to power the loads as
Decoupling capacitors such as 0.1 μF or 0.22 μF should described in the reference manual or data sheet. Do not
also be used. connect any loads to these LDO outputs:
NVCC_PLL_OUT or VDDUSB_CAP. VDDHIGH_CAP is
restricted to MX6 loads.
5. One 0.22 μF decoupling capacitor should be If the nominal value is larger than recommended,
connected to VDD_SNVS_CAP, an on-chip LDO power-up/down ramp time is excessive and
regulator output. A bulk capacitor is not necessary. suspend/resume operation cannot be guaranteed. Select
a small capacitor with low ESR.
Note: Do not connect any loads to VDD_SNVS_CAP.
6. Maximum ripple voltage requirements. Common requirement for ripple noise should be less than
5% Vp-p of supply voltage average value.
Related power rails affected: all VDD_xxx_IN and
VDD_xxx_CAP.
7. NVCC_LVDS2P5 must be powered-on even when The DDR pre-drivers share the NVCC_LVDS2P5 power
not using the LVDS interface. rail with the LVDS interface. VDDHIGH_CAP can be
utilized as the power source; tie NVCC_LVDS2P5 to
VDDHIGH_CAP.
8. Account for the different power design on • i.MX 6Quad and 6Dual chips can support three different
NVCC_EIM between i.MX 6Quad and 6Dual chips and EIM power rail voltage levels: NVCC_EIM0(K19),
i.MX 6DualLite and 6Solo chips. NVCC_EIM1(L19), and NVCC_EIM2(M19).
• i.MX 6DualLite and 6Solo chips support one EIM power
rail: NVCC_EIM (K19, L19, M19). The three power
contacts must be connected to same power supply.
9. If VDD_SNVS_IN is directly supplied by a coin cell, a When no power is supplied to VDD_VSNVS_IN, the diode
schottky diode is required between VDD_HIGH_IN and limits the voltage difference between the two on-chip
VDD_SNVS_IN. The cathode is connected to SNVS power domains to approximately 0.3 V. The
VDD_SNVS_IN. processor is designed to allow current flow between the
Alternately, VDD_HIGH_IN and VDD_SNVS_IN can be two SNVS power domains proportional to the voltage
tied together if the real-time clock function is not difference.
needed during system power-down.
Design Checklist
Table 2-7. Oscillator and clock recommendations
1. Precision 32.768 kHz oscillator The capacitors implemented on either side of the crystal
Connect a crystal between RTC_XTALI and are about twice the crystal load capacitance. To hit the
RTC_XTALO. Choose a crystal with a maximum of target oscillation frequency, board capacitors need to be
100 kΩ ESR (equivalent series resistance) and follow reduced to compensate for board and chip parasitic
the manufacturer’s recommendation for loading capacitance; typically 15–16 pF is employed.
capacitance. The integrated oscillation amplifier has an on-chip
Do not use an external biasing resistor because the self-biasing scheme, but is high-impedance (relatively
bias circuit is on-chip. weak) to minimize power consumption. Care must be
taken to limit parasitic leakage from RTC_XTALI and
RTC_XTALO to either power or ground (> 100 MΩ) as
this negatively affects the amplifier bias and causes a
reduction of startup margin.
Use short traces between the crystal and the processor,
with a ground plane under the crystal, load capacitors,
and associated traces.
2. External kilohertz source The voltage level of this driving clock should not exceed
If feeding an external clock into the device, the voltage of VDD_SNVS_CAP and the frequency
RTC_XTALI can be driven DC-coupled with should be <100 kHz under typical conditions. Do not
RTC_XTALO floated or driven with a complimentary exceed VDD_SNVS_CAP or damage/malfunction may
signal. occur. The RTC_XTALI signal should not be driven if the
VDD_SNVS_CAP supply is off. This can lead to damage
or malfunction.
For RTC_XTALI VIL and VIH voltage levels, see the latest
i.MX 6 series datasheet available at www.freescale.com.
Note that if this external clock is stopped, the internal ring
oscillator starts automatically.
3. Loose-tolerance 40 kHz oscillator When a high-accuracy real-time clock is not required, the
An on-chip loose-tolerance ring oscillator is available system may use the on-chip 40 kHz oscillator. The
of approximately 40 kHz. If RTC_XTALI is tied to GND tolerance is ± 50%.
and RTC_XTALO is floating, the on-chip oscillator is The ring oscillator starts faster than an external crystal
automatically engaged. and is used until the external crystal reaches stable
oscillation. The ring oscillator also starts automatically if
no clock is detected at RTC_XTALI at any time.
4. Precision 24 MHz oscillator Freescale BSP software requires 24 MHz on this clock.
Connect a fundamental-mode crystal between XTALI This clock is used as a reference for USB, PCIe, and
and XTALO. An 80 Ω typical ESR crystal rated for a SATA, so there are strict frequency tolerance and jitter
maximum drive level of 250 μW is acceptable. requirements. See Table 2-20 for guidelines. See the
Alternately, a 50 Ω typical ESR crystal rated for a crystal oscillator (XTALOSC) reference manual chapter
maximum drive level of 200 μW may be used. See the and relevant interface specification chapters for details.
engineering bulletin EB830 on www.freescale.com for To access a calculator for the 24 MHz crystal drive level,
additional options. see EB830 on the i.MX Community.
5. External megahertz source For XTALI VIL and VIH voltage levels, see the latest i.MX
If feeding an external clock into the device, XTALI can 6 series datasheet. This clock is used as a reference for
be driven DC-coupled with XTALO floated. USB, PCIe, and SATA, so there are strict frequency
tolerance and jitter requirements. See Table 2-20 for
guidelines. See the crystal oscillator (XTALOSC)
reference manual chapter and relevant interface
specification chapters for details.
Design Checklist
Table 2-7. Oscillator and clock recommendations (continued)
6. CLK1_P/CLK1_N and CLK2_P/CLK2_N are LVDS The clock inputs/outputs are general-purpose differential
input/output differential pairs compatible with high-speed clock Input/outputs.
TIA/EIA-644 standard. The frequency range is 0 to Any or both of them can be configured:
600 MHz. • As inputs to feed external reference clocks to the
Alternatively, a single-ended signal can be used to on-chip PLLs and/or modules, for example as
drive a CLKx_P input. In this case, the corresponding alternate reference clock for PCIe or/and SATA or
CLKx_N input should be tied to a constant voltage video/audio interfaces.
level equal to 50% of VDD_HIGH_CAP. Termination • As outputs to be used as either a reference clock or as
should be provided with high-frequency signals. a functional clock for peripherals, for example an
See the LVDS pad electrical specification in the data output of the PCIe master clock (root complex use).
sheet for further details. See the chip reference manual for details on the
After initialization, the CLKx inputs/outputs can be respective clock trees.
disabled (if not used) by the PMU_MISC1 register. If
unused, any or both of the CLKx_N/P pairs may be left
floating.
7. Bias XTALI with a 2.2 MΩ resistor to GND. Mount The XTALI bias must be adjusted externally to ensure
the resistor close to the XTALI ball. reasonable start-up time. Without the resistor, start-up
time may be 200 ms or more.
Table 2-8. Reset and ONOFF recommendations
1. The POR_B input must be asserted low at power-up A reset switch may be wired to the chip’s POR_B, which
and remain asserted until after the last power rail for is a cold-reset negative-logic input that resets all modules
devices required for system boot are at their working and logic in the IC. POR_B may be used in addition to
voltage. internally generated power-on reset signal (logical AND,
both internal and external signals are considered active
low).
2. For portable applications, the ONOFF input may be A brief connection to GND in OFF mode causes the
connected to an ON/OFF SPST push-button switch. internal power management state machine to change
On-chip debouncing is provided, and this input has an state to ON.
on-chip pullup. In ON mode, a brief connection to GND generates an
If not used, ONOFF should be a no connect. interrupt (intended to be a software-controllable
power-down).
An approximate 5 second or more connection to GND
causes a forced OFF.
Design Checklist
Table 2-9. Gigabit Ethernet Recommendations
1. This chip requires a 125 MHz reference clock If NVCC_ENET is powered at 3.3 V, the minimum VIH
feeding the ENET_REF_CLK input. This reference level is 70% of 3.3 V or 2.3 V. Designers should ensure
clock should be sourced from an external 125 MHz that there is margin to this minimum value. A starting value
oscillator or an external PHY. could be 500 mV margin, resulting in a requirement of
Designers should be aware of the 125 MHz reference 2.8 V for the logic high. See the Freescale Smart Devices
output level of the PHY because ENET_REF_CLK is development designs for a suitable low-cost level
on the NVCC_ENET supply rail, not the NVCC_RGMII translator.
rail.
2. For IEEE-1588 operation, the GPIO_16 ball must be This allows use of time stamping on the RGMII interface.
configured as either one of the following:
• A no connect to allow the internal time stamp clock
to route through its IOMUX cell to the RGMII
interface
• Driven by an external clock source for the time
stamp
Table 2-10. PCIe recommendations
1. Termination is required on the differential clock lines. These termination resistors should be placed as close as
Connect two 49.9 Ω resistors, one between REFCLK- possible to the receiver device inputs in case the chip
and GND, the other between REFCLK+ and GND. LVDS clock outputs are used as the REFCLK source for
Alternately, Connect a 100 Ω resistor between the PCIe endpoint device.
REFCLK- and REFCLK+.
2. The differential transmitter must be ac coupled. Use To ensure PCIe specification compliance, ac coupling is
a 0.1 uF-series capacitor on PCIE_TXP and a second required at each transmitter. The receiver must be dc
0.1 uF on PCIE_TXM. coupled.
Table 2-11. HDMI recommendations
1. The designer must ensure that a suitable level shifter The i.MX 6 processors’ I2C cannot operate at the 5 V
and driver be used to interface the chip’s I2C with the required by HMDI EDID. The i.MX 6 processors’ supply
HDMI monitor. limit is 3.6 V maximum.
In addition, ESD (electrostatic discharge) protection The designer could consider the ON Semiconductor
must be used on all HDMI single-ended and differential CM2020 for ESD protection and I2C level conversion.
signals mounted near the board’s HDMI connector. Note: Freescale cannot recommend one supplier over
another and does not suggest that this is the only
HDMI interface chip supplier.
2. DDC (EDID) must be on a dedicated I2C When HDCP is enabled, a dedicated I2C is controlled by
(DDC_SCL/DDC_SDA) port when HDCP the HDMI PHY to exchange the HDCP encryption key and
(High-Bandwidth Digital Content Protection) is enabled. must sync several times per second. DDC does not
behave like a common I2C and cannot be controlled by the
ARM® CPU with HDCP enabled.
Design Checklist
Table 2-12. USB recommendations
1. USB OTG The processor should turn VBUS on as required.

To comply with the USB OTG specification, the VBUS
supply on the OTG connector should default to off when
the boards power up.
2. USB Host Tie USB_H1_VBUS to an unswitched 5 V supply for the

USB_H1_VBUS should be directly connected to a 5 V typical use case. However, if the your system is a USB
supply. device, then USB_H1_VBUS may be a no connect.
Table 2-13. Reference resistor recommendations
1. HDMI_REF – Connect an external 1.6 kΩ 1% If HDMI is unused, the reference resistor may be
resistor to GND. populated if desired for manufacturability purposes, or left
no-connect for cost savings.
2. SATA_REXT – Connect an external 191 Ω 1% The impedance calibration process requires connection of
resistor to GND. this reference resistor.
If SATA is unused, the reference resistor may be
populated if desired for manufacturability purposes, or left
3. PCIE_REXT – Connect an external 200 Ω 1% The impedance calibration process requires connection of
resistor to GND. this reference resistor.
If PCIe is unused, the reference resistor may be populated
if desired for manufacturability purposes, or left
4. CSI_REXT – Connect an external 6.04 kΩ 1% If CSI is unused, the reference resistor may be populated
resistor to GND. if desired for manufacturability purposes, or left
5. DSI_REXT – Connect an external 6.04 kΩ 1% If DSI is unused, the reference resistor may be populated
resistor to GND. if desired for manufacturability purposes, or left
Table 2-14. Miscellaneous recommendations
1. The TEST_MODE input is internally connected to an This input is reserved for Freescale manufacturing use.
on-chip pulldown device. The user can either float this
signal or tie it to GND.
2. For termination of unused analog interfaces, see —

Table 2-21.
3. VDD_FA and FA_ANA should be tied to GND. These inputs are reserved for Freescale manufacturing
use. Best practice is to tie them to ground to avoid floating
inputs.
Design Checklist
Table 2-14. Miscellaneous recommendations (continued)
4. GPANAIO must be a no connect. This output is reserved for Freescale manufacturing use.
5. NC contacts are no connect and should be floated. Depending on the feature set, some versions of the IC
may have NC contacts connected inside the BGA.
2.3 Bus isolation circuit

The following figure provides supporting information for Table 2-2, recommendation #1
Figure 2-1. Boot configuration for development mode
2.4 DDR reference circuit

The following table is a resistor chart (see Table 2-1 recommendation #2). The recommendations are
appropriate for designs with DDR memory chips with a maximum Vref input current of 2µA each.
Design Checklist
Table 2-15. DDR Vref resistor sizing guideline
Number of DRAM with 2 μA Vref Resistor divider value

input current (2 resistors)
2 ≤1.21 kΩ 1%
2 ≤1.54 kΩ 0.5%
2 ≤2.32 kΩ 0.1%
4 ≤768 Ω 1%
4 ≤1 kΩ 0.5%
4 ≤1.5 kΩ 0.1%
2.5 I2C address

The following table shows a spreadsheet approach to avoid I2C conflicts as referenced in Table 2-4
recommendation #2.
NOTE
The example in this section only applies to the Freescale reference design
board.
Notice that although there are no slave address conflicts, the shaded cell in the table highlights a potential
bus speed issue as discussed in Table 2-4. The AM-FM tuner limits the maximum bus rate to 250 kbps,
but the bus data rate cannot exceed the slowest peripheral on the bus.
If the system cannot tolerate the 250 kbps rate for proper operation, the AM-FM tuner must be moved to
another I2C port. If the I2C bus rate exceeded the AM-FM tuner module’s maximum bus rate, the I2C bus
operation might fail or become unpredictable. The slow peripheral may unpredictably take over the bus or
might malfunction in some other way.
Table 2-16. I2C bus example spreadsheet
Slave addresses supported on

Selected system
Peripheral Bus activity level Speed (kbps) the peripheral
address (hex)
(hex)
PMIC Low 400 68 68
Port Expander Low 400 30, 32, 34 30
AM-FM Tuner Med 250 C0, C2, C4, C6 C0
A/D Converter Med 400 40, 42 40
Audio CODEC Low 400 90, 92, 94, 96 90
Design Checklist
Assuming the system can function properly with a reduced bus rate of 250 kbps, the following table
provides a possible I2C port usage table.
Table 2-17. I2C port usage scenario
i.MX6 I2C ports Ball name Function Speed (kbps)
Port 1
Port 1
Port 2 KEY_ROW3 I2C2_SDA 250
Port 2 EIM_EB2 I2C2_SCL 250
Port 3
Port 3
2.5.1 I2C clock speed and division factors (IFDR)

The I2C clock is sourced from PERCLK_CLK_ROOT which is routed from IPG_CLK_ROOT. The I2C
clock frequency can be easily obtained using the following formula:
I2C clock Frequency = (PERCLK_ROOT frequency)/(division factor corresponding to IFDR)
By default, the IPG_CLK_ROOT and PERCLK_CLK_ROOT frequencies are set to 49.5MHz, where the
root clock is sourced from PLL2’s PFD2. Obtaining the frequencies can be accomplished using the
following:
PLL2 = 528MHz
PLL2_PFD2 = 528MHz * 18 / 24 = 396MHz
IPG_CLK_ROOT = (PLL2_PFD2 / ahb_podf )/ ipg_podf = (396MHz/4)/2 = 49.5MHz
PER_CLK_ROOT = IPG_CLK_ROOT/perclk_podf = 49.5MHz/1 = 49.5MHz
NOTE
The above calculation assumes that the default CCM register settings,
routing, and division factors are used. If different routing, PFD values,
and/or division factors are used, the user must adjust the parameters
accordingly to calculate the correct clock frequency.
IFDR, division factor and resulting I2C CLK frequencies are indicated in the table below. Resulting
frequencies will vary according to the PERCLK_CLK_ROOT frequencies selected.
Design Checklist
Table 2-18 assumes PERCLK_CLK_ROOT = 49.5MHz.

Table 2-18. IFDR1
IFDR Division factor Frequency (kHz)
0 30 1650
1 32 1546.875
2 36 1375
3 42 1178.571
4 48 1031.25
5 52 951.9231
6 60 825
7 72 687.5
8 80 618.75
9 88 562.5
A 104 475.9615
B 128 386.7188
C 144 343.75
D 160 309.375
E 192 257.8125
F 240 206.25
10 288 171.875
11 320 154.6875
12 384 128.9063
13 480 103.125
14 576 85.9375
15 640 77.34375
16 768 64.45313
17 960 51.5625
18 1152 42.96875
19 1280 38.67188
1A 1536 32.22656
1B 1920 25.78125
1C 2304 21.48438
1D 2560 19.33594
1E 3072 16.11328
1F 3840 12.89063
Design Checklist
Table 2-18. IFDR1 (continued)
20 22 2250
21 24 2062.5
22 26 1903.846
23 28 1767.857
24 32 1546.875
25 36 1375
26 40 1237.5
27 44 1125
28 48 1031.25
29 56 883.9286
2A 64 773.4375
2B 72 687.5
2C 80 618.75
2D 96 515.625
2E 112 441.9643
2F 128 386.7188
30 160 309.375
31 192 257.8125
32 224 220.9821
33 256 193.3594
34 320 154.6875
35 384 128.9063
36 448 110.4911
37 512 96.67969
38 640 77.34375
39 768 64.45313
3A 896 55.24554
3B 1024 48.33984
3C 1280 38.67188
3D 1536 32.22656
3E 1792 27.62277
3F 2048 24.16992
1
Shaded cells indicate frequency is outside of the range that guarantees operation.
Design Checklist
2.6 JTAG signal termination

The following table is a JTAG termination chart (see recommendations in Table 2-5).
Table 2-19. JTAG interface summary
JTAG signal I/O type On-chip termination External termination
JTAG_TCK Input 47 kΩ pullup Not required;

can use 10 kΩ pullup
JTAG_TMS Input 47 kΩ pullup Not required;

JTAG_TDI Input 47 kΩ pullup Not required;

JTAG_TDO 3-state output Keeper Do not use pullup or pulldown
JTAG_TRSTB Input 47 kΩ pullup Not required;

JTAG_MOD Input 100 kΩ pullup Use 1 kΩ pulldown or tie to GND
2.7 Oscillator tolerance

The following table provides 24 MHz oscillator tolerance guidelines (see Table 2-7, recommendations #4
and #5). Because these are guidelines, the designer must verify all tolerances per the official specifications.
Table 2-20. 24 MHz crystal tolerance guidelines
Tolerance
Interface
(± ppm)
Ethernet 50
HDMI 100
SATA 350
USB2.0 500
PCIe 300
2.8 Unused analog interfaces

Table 2-21 shows the recommended connections for unused analog interfaces (see Table 2-14,
recommendation #2).
Design Checklist
Table 2-21. Recommended connections for unused analog interfaces
Module Contact name Recommendations if unused
CCM CLK1_N, CLK1_P, CLK2_N, CLK2_P Float
CSI CSI_CLK0M, CSI_CLK0P, CSI_D0M, CSI_D0P, CSI_D1M, CSI_D1P, CSI_D2M, Float

CSI_D2P, CSI_D3M, CSI_D3P, CSI_REXT
DSI DSI_CLK0M, DSI_CLK0P, DSI_D0M, DSI_D0P, DSI_D1M, DSI_D1P, Float

DSI_REXT
HDMI HDMI_CLKM, HDMI_CLKP, HDMI_D0M, HDMI_D0P, HDMI_D1M, HDMI_D1P, Float
HDMI_D2M, HDMI_D2P, HDMI_DDCEC, HDMI_HPD, HDMI_REF
HDMI_VP, HDMI_VPH Ground
LDB LVDS0_CLK_N, LVDS0_CLK_P, LVDS0_TX0_N, LVDS0_TX0_P, Float

LVDS0_TX1_N, LVDS0_TX1_P, LVDS0_TX2_N, LVDS0_TX2_P,
LVDS0_TX3_N, LVDS0_TX3_P, LVDS1_CLK_N, LVDS1_CLK_P,
LVDS1_TX0_N, LVDS1_TX0_P, LVDS1_TX1_N, LVDS1_TX1_P,
LVDS1_TX2_N, LVDS1_TX2_P, LVDS1_TX3_N, LVDS1_TX3_P
MLB MLB_CN, MLB_CP, MLB_DN, MLB_DP, MLB_SN, MLB_SP Float
PCIe PCIE_REXT, PCIE_RXM, PCIE_RXP, PCIE_TXM, PCIE_TXP Float
PCIE_VP, PCIE_VPH, PCIE_VPTX Ground1
RGMII RGMII_RD0, RGMII_RD1, RGMII_RD2, RGMII_RD3, RGMII_RX_CTL, Float

RGMII_RXC, RGMII_TD0, RGMII_TD1, RGMII_TD2, RGMII_TD3,
RGMII_TX_CTL, RGMII_TXC
SATA SATA_REXT, SATA_RXM, SATA_RXP, SATA_TXM, SATA_TXP Float
SATA_VP, SATA_VPH Ground1
USB USB_H1_DN, USB_H1_DP, USB_H1_VBUS, USB_OTG_CHD_B, Float

USB_OTG_DN, USB_OTG_DP, USB_OTG_VBUS
1 These supplies must remain powered if boundary scan test needs to be done
Design Checklist
Chapter 3
This chapter provides recommendations to assist design engineers with the correct layout of their i.MX 6
series-based system. The majority of the chapter discusses the implementation of the DDR interface, but
it also provides recommendation for power, the HDMI, SATA, LVDS, PCIe, USB, reference resistors, ESD
and related emissions.
This chapter uses the i.MX6DQ SABRE SD board as its reference for illustrating the key concepts. See
the i.MX6DQ SABRE SD board layout files as a companion to this chapter.
3.1 Basic design recommendations

The i.MX 6Dual/6Quad processor comes in a 21 × 21 mm package with 0.8 mm ball pitch. The ball-grid
array contains 25 rows and 25 columns, making it a 624 ball BGA package. For detailed information
about the package, see the i.MX 6 series Consumer and Automotive datasheets.
The following figure shows the ball-grid array. Figure 3-2 shows additional package information.
Figure 3-1. i.MX 6DQ/SDL ball-grid array
Figure 3-2. i.MX 6DQ/6SDL package information
It is critical to maintain the recommended footprint of a 16 mils pad with a 20 mil open solder mask for
ease of fanout. In this case, the solder paste is the same as the pad with 16 mil, which allows an air gap of
15.496-mil between pads.
When using the Allegro tool, optimal practice is to use the footprint as created by Freescale. When not
using the Allegro tool, use the Allegro footprint export feature (supported by many tools). If export is not
possible, create the footprint as per the package mechanical dimensions outlined in the product data sheet.
3.1.1 Fanout illustrations

The following figures show the top and bottom layer fanouts for the i.MX 6Dual/6Quad chip.
Figure 3-3. i.MX6DQ fanout example, top layer view
Figure 3-4. i.MX6DQ fanout example, bottom view
The colors signify the following:

• Top layer
— Red = etch
— Yellow = pad
— Gray = vias
• Bottom layer
— Cyan = GND net
— Brown = power rails
3.1.2 Placing decoupling capacitors

The fanout scheme creates a four quadrant structure that facilitates the placement of decoupling bulk
capacitors on the bottom side of the PCB.
The 0201 decoupling and 0603 bulk capacitors should be mounted as close as possible to the power vias.
The distance should be less than 50 mils. Additional bulk capacitors can be placed near the edge of the
BGA via array. Placing the decoupling capacitors close to the power balls is critical to minimize inductance
and ensure high-speed transient current demand by the processor.
A correct via size is critical for preserving adequate routing space. The recommended geometry for the via
pads is: pad size 18 mils and drill 8 mils.
The following list provides the main recommendations for choosing the correct decoupling scheme for the
i.MX6 family boards.
• Place the largest capacitance in the smallest package that budget and manufacturing can support.
• For high speed bypassing, select the required capacitance with the smallest package (for example,
0.22 μF and package 0201).
• Minimize trace length (inductance) to small caps.
• Series inductance cancels out capacitance.
• Tie caps to GND plane directly with a via.
• Place capacitors close to the power contact of the associate package designed from the schematic.
The i.MX6 SABRE SD (Smart Devices) CPU uses the preferred BGA power decoupling design. Note that
the layout is available through www.freescale.com. Customers should use the reference design strategy for
power and decoupling.
3.2 Stackup recommendations

High-speed design requires a good stackup in order have the right impedance for the critical traces. The
constraints for the trace width may depend on a number of factors, such as the board stackup and associated
dielectric and copper thickness, required impedance, and required current (for power traces). The
Freescale reference design uses a minimum trace width of 3 mils for the DDR routing. The stackup also
determines the constraints for routing and spacing.
Consider the following when designing the stackup and selecting the material for your board.
• Board stack-up is critical for high-speed signal quality.
• You must preplan impedance of critical traces
• High-speed signals must have reference planes on adjacent layers to minimize cross-talk.
• FSL reference design equals Isola 370HR.
• FSL validation boards equals Isola FR408.

The recommended stackup is 8-layers, with the layer stack as shown in the following figure. The lefthand
image shows the detail provided by Freescale inside the fabrication detail as a part of the Gerber files. The
righthand side shows the solution suggested by the PCB fabrication company for our requirements.
Figure 3-5. Layer stack SABRE SD board

1 Additional power planes to support i.MX 6Dual/6Quad and i.MX 6Solo power options only.
The following table shows a working stack-up implementation:

Table 3-1. Stackup implementation
Single ended Differential

Layers
Trace width Impedance Trace width Trace spacing Impedance Trace width Trace spacing Impedance
(Mils) (Ωs) (Mils) ‘Airgap’ (Mils) (Ωs) (Mils) ‘Airgap’ (Mils) (Ωs)
TOP 4.7 50 4.3 5.7 90 3.7 5.3 100
INT1 4.5 50 4.2 5.8 90 3.8 5.2
INT2 4.5 50 4.2 5.8 90 3.8 5.2
BOT 4.7 50 4.3 5.7 90 3.7 5.3 100
Figure 3-6. Example top layer impedance solution from PCB fabricator
3.3 DDR connection information

The following figures show the block diagrams from the reference design boards for the DDR3 interface
and the LPDDR2 interface (respectively) with the i.MX6DQ/SDL.
Figure 3-7. Connection between i.MX6DQ/SDL and DDR3
Figure 3-8. Connection between i.MX6DQ/SDL and LPDDR2
The DDR3 interface is one of the most critical interfaces for chip routing. It must have the controlled
impedance for the single ended traces be equal to 50 Ω and for the differential pairs be equal to 100 Ω.
The following figure shows the physical connection scheme for both top and bottom placement of the DDR
chips, showing the final placement of the DDR3 memory and the decoupling capacitors. The blue figure
shows the top layer and the red figure shows the bottom layer. It is very important to place the memory as
close to the processor as possible to reduce trace capacitance and keep the propagation delay to the
minimum. Follow the reference board layout as a guideline for memory placement and routing.

Figure 3-9. Final placement of memories and decoupling capacitors
3.4 DDR routing rules

DDR3 routing can be accomplished in two different ways: routing all signals at the same length or routing
by byte group.
Routing all signals at the same length can be more difficult at first because of the tight space between the
DDR and the processor and the large number of required interconnects. However, it is the better way
because it makes signal timing analysis straightforward. Ideally, we could route all the signals at the same
length, However, it could be difficult because of the large number of connections in the tight space between
the DDR and the processor. The following table explains the rules for routing the signals by the same
length.
Table 3-2. DDR3 routing by the same length
Signals Total length Recommendations
Address and Bank Clock length Match the signals ±25 mils of the value
specified in the length column
Data and Buffer Clock length
Control signals Clock length
Clock Longest trace ≤ 3 inches Match the signals of clocks signals

DRAM_SDCLK[1:0] ±5 mils. Each differential clock pair
DRAM_SDQS[7:0] and Clock length Match the signals of DQS signals

DRAM_SDQS[7:0]_B ±10 mils of the value specified in the
length column.
Routing by byte group requires better control of the signals of each group. It is also more difficult for
analysis and constraint settings. However, its advantage is that the constraint to match lengths can be
applied to a smaller group of signals. This is often more achievable once the constraints are properly set.
The following table explains the rules for routing the signals by byte group.
Table 3-3. DDR3 routing by byte group
Length
Chip signals Group Recommendations
Min Max
DRAM_SDCLK[1:0] Clock Short as possible 2.25 inches Match the signals ± 5 mils.
DRAM_SDCLK_B[1:0] 2.25 inches is recommended.
DRAM_A[15:0] Address Clock (min) – 200 Clock (min)1 Match the signals ± 25 mils.
DRAM_SDBA[2:0] and Command
DRAM_RAS DRAM_CAS
DRAM_SDWE
Table 3-3. DDR3 routing by byte group (continued)
Length
Chip signals Group Recommendations
Min Max
DRAM_D[7:0] Byte Group 1 — Clock (min) Match the signals of each byte group ± 25
DRAM_DQM0 mils.
DRAM_SDQS0 Match the differential signals of DQS ± 10
DRAM_SDQS0_B mils.
DRAM_D[15:8] Byte Group 2 — Clock (min)

DRAM_DQM1
DRAM_SDQS1
DRAM_SDQS1_B

DRAM_DQM2
DRAM_SDQS2
DRAM_SDQS2_B

DRAM_DQM3
DRAM_SDQS3
DRAM_SDQS3_B

DRAM_DQM4
DRAM_SDQS4
DRAM_SDQS4_B

DRAM_DQM5
DRAM_SDQS5
DRAM_SDQS5_B

DRAM_DQM6
DRAM_SDQS6
DRAM_SDQS6_B

DRAM_DQM7
DRAM_SDQS7
DRAM_SDQS7_B
DRAM_CS[1:0] Control signals Clock (min) – 200 Clock (min) Match the signals ± 50 mils.
DRAM_SDCKE[1:0]
DRAM_SDODT[1:0]
1. Clock (min)—The shortest length of the clock group signals because this group has a ± 5 mil matching tolerance.
Finally, the impedance for the signals should be 50 Ω for single ended and 100 Ω for differential pairs.
3.5 Routing considerations

The chip can handle up to 4 Gbytes of DRAM memory. i.MX6 DDR routing needs to be separated into
three groups: data, address, and control. Each group has its own method of routing from an i.MX 6
serieschip to DDR memory. The DDR layout has 2 Gbyte and 4 Gbyte options.
3.5.1 Swapping data lines

The DDR3 pin swapping technique for the data bus lines within bytes makes it easier to:
• Route direct lines
• Avoid changes between layers
The rules are as follows:
• Hardware write leveling – lowest order bit within byte lane must remain on lowest order bit of lane
by JEDEC compliance (see the “Write Leveling” section in JESD79-3E)
— D0, D8, D16, D24, D32, D40, D48, and D56 are fixed
— Other data lines free to swap within byte lane
• JEDEC DDR3 memory restrictions are:
– No restrictions for complete byte lane swapping
– DQS and DQM must follow lanes
NOTE
If byte lane swapping was done, target DDR IC register read value must be
transposed according to the data line swapping.
3.5.2 DDR3 (64 bits) T topology considerations

Be sure to take into account the following when designing a T-topology system.
• Follow the routing rules described in Table 3-3.
• Termination resistors not required.
• Short routing lengths and on-chip drive strength control.
• Your design is limited to 4 DDR chips.
• DDR3, 2 GBytes using latest memories (4 GBytes coming).
3.5.3 DDR3 (64 bits) Fly-by topology considerations

Pay attention to the following recommendations when the Fly-by topology and routing technique.
• DDR controller provides address mirroring when using two chip selects, which aids address line
routing for memories on both sides of board.
• Bus termination resistors are required.
3.5.4 2-Gigabyte recommendations

The 2 Gbyte option has four memories. You should follow these recommendations for best practice:
• Have a balanced routing for the T connection.
• Avoid having many layer transitions.
• Do not cross split reference planes during the routing.
The following figure shows the topology for the ADDR/CMD/CTRL signals. It has a tree topology. Note
the balanced T routing.
Figure 3-10. ADDR/CMD/CTRL signal topology
The routing for the data groups depends on the bus size. The following figure shows the point-to-point data
bus connection, with routing by byte group.
Figure 3-11. Point-to-point data bus connection (routing by byte group)
NOTE
i.MX 6Solo only uses the first two pairs of the 2 Bytes groups. All others are
disabled.
3.5.5 4-Gigabyte recommendations

The following diagrams show the 4 Gbyte recommendations using both chip selects (CS[1:0]) and loading
2 GBytes to each one. This option has eight memories and requires the addition of a termination resistor.
Route the ADDR/CMD signals as shown in the following figure.
Figure 3-12. ADDR/CMD signal topology
Figure 3-13. CTRL signal topology
Figure 3-14. Data bus routing topology
Figure 3-15. Clock routing topology
3.5.6 Four chips T topology routing examples

The figures in this section show examples for the routing of the 2 GByte DDR3 memories. Figure 3-16
through Figure 3-18 are a guideline of the T configuration routing with eight layers PCB.Table 3-4 shows
the color coding used in the figures.
Table 3-4. Color code
Color Meaning
Soft Green ADD & CMD Signals
Yellow Clocks
Soft Pink Data Byte Group 0
Purple Data Byte Group 1
Blue Data Byte Group 2
Brown Data Byte Group 3
Orange Data Byte Group 4
Green Data Byte Group 5
Olive Green Data Byte Group 6
Soft Brown Data Byte Group 7
Gray DDR_1V5 & DDR_VREF
Soft Red Control Signals
Figure 3-16. Top layer DDR3 routing
Figure 3-17. Internal L6 DDR3 routing
Figure 3-18. Bottom layer DDR3 routing
The following table shows the total etch of the signals for the byte 0 and byte 1 groups. The layout is an
example, using 2000 mils for the clock.
Table 3-5. Total signal etch (DDR3)
Signals Length (Mils)
DRAM_D0 1025.349
DRAM_D1 1028.996
DRAM_D2 1028.752
DRAM_D3 1021.158
DRAM_D4 1021.930
DRAM_D5 1025.398
DRAM_D6 1025.564
DRAM_D7 1029.326
Table 3-5. Total signal etch (DDR3) (continued)
DRAM_DQM0 1028.555
DRAM_SDQS0 1023.419
DRAM_SDQS0_B 1023.373
DRAM_D8 648.862
DRAM_D9 654.371
DRAM_D10 652.653
DRAM_D11 653.712
DRAM_D12 650.961
DRAM_D13 648.433
DRAM_D14 649.588
DRAM_D15 651.781
DRAM_DQM1 653.106
DRAM_SDQS1 669.240
DRAM_SDCLK0 2120.044
DRAM_SDCLK0_B 2118.283
DRAM_SDCLK1 2112.518
3.5.7 Eight chips fly-by topology routing examples

The figures in this section show examples for the routing of 4-Gbyte DDR memories. These figures are a
guideline of the routing by layer using the fly by configuration topology. They use the same color code
shown in Table 3-4.
NOTE
The SABRE SD board referenced in the beginning of this chapter does not
use eight DDR chips. The following screen shots are from the validation
board layout.
Figure 3-19. Top DDR3 routing
Figure 3-24. Bottom DDR3 routing
The following table shows the total etch of the signals for the byte 0 and byte 1 groups.
Table 3-6. Total signal etch (DDR3)
DRAM_D0 1244.97
DRAM_D1 1252.82
DRAM_D2 1237.48
DRAM_D3 1242.95
DRAM_D4 1240.12
DRAM_D5 1254.37
DRAM_D6 1254.58
DRAM_D7 1238.18
DRAM_DQM0 1297.45
DRAM_SDQS0 1295.34
DRAM_D8 1103.69
DRAM_D9 1116.14
DRAM_D10 1105.01
DRAM_D11 1105.17
DRAM_D12 1120.4
DRAM_D13 1123.06
DRAM_D14 1105.72
DRAM_D15 1111.24
DRAM_DQM1 1152.16
DRAM_SDQS1 1158.48
DRAM_SDCLK0 4723.96
DRAM_SDCLK1 4750.69
3.5.8 High speed signal routing recommendations

The following list provides recommendations for routing traces for high speed signals. Note that the
propagation delay and the impedance control should match in order to have the correct communication
with the devices.
• High-speed signals (DDR, RGMII, display) must not cross gaps in the reference plane.
• Avoid creating slots, voids, and splits in reference planes. Review via voids to ensure they do not
create splits (space out vias).
• A solid GND plane must be directly under crystal, associated components, and traces.
• Clocks or strobes that are on the same layer need at least 2.5× spacing from an adjacent trace (2.5×
height from reference plane) to reduce cross-talk.
• All synchronous modules should have bus length matching and relative clock length control.
— For SD module interfaces:
– Match data and CMD trace lengths (length delta depends on bus rates)
– CLK should be longer than the longest signal in the Data/CMD group (+5 mils)
— Similar DDR rules must be followed for data, address and control as for SD module interfaces.
3.5.9 Ground plane recommendations

This section provides examples of good practices and how to avoid common user mistakes when flowing
the ground planes layers.
The following two figures show common examples of poor GND planes. The copper plane is represented
by the color gray in Figure 3-25 and by the horizontal green lines in Figure 3-26.
Figure 3-25. Poor GND plane 1
Figure 3-26. Poor GND plane 2
Spacing the vias some mils apart facilitates the GND copper flowing in the plane. The following figures
show good practices of ground planes.
Figure 3-27. Good layout GND plane detail
Figure 3-28. Good layout GND plane detail
3.6 DDR power recommendations

The following recommendations apply to the VREF (P0V75_REFDDR) voltage reference plane.
• Use 30 mils trace between decoupling cap and destination.
• Maintain a 25 mils clearance from other nets.
• Isolate VREF and/or shield with ground.
• Decouple using distributed 0.22 μF capacitors by the regulator, controller, and devices.
• Place one 1.0 μF near the source of VREF: one near the VREF pin on the controller and two
between the controller and the devices.
The following recommendations apply to the VTT (DDR_VTT) voltage reference plane. The figures are
examples from the evaluation board for the VTT reference schematic.
• Place the VTT island on the component side layer at the end of the bus behind the DRAM devices.
• Use a wide-island trace for current capacity.
• Place the VTT generator as close to termination resistors as possible to minimize impedance
(inductance).
• Place one or two 0.1 μF decoupling capacitors by each termination RPACK on the VTT island to
minimize the noise on VTT. Other bulk (10–22 pF) decoupling is also recommended to be placed
on the VTT island.
Figure 3-29. DDR_VTT evaluation board example
Figure 3-30. DDR_VTT evaluation board examples
3.7 PCI Express interface recommendations

This chip provides a ×1 PCIe lane. The PCIe module supports PCI Express Gen 2.0 interfaces at 5 Gb/s.
It is also backwards compatible to Gen 1.1 interfaces at 2.5 Gb/s.
NOTE
Lane ×1 is composed of two differential signals pairs: one TXD signal pair
and one RXD signal pair.
Table 3-7. PCI Express signal descriptions
Signal name Signal group Description
PCIE_TXP, PCIE_TXM Data PCI Express transmit differential pair
PCIE_RXP, PCIE_RXM Data PCI Express receive differential pair
3.7.1 PCI Express general routing guidelines

Use the following recommendations for PCI Express general routing:
• The trace width and spacing of the lanes ×1 signals should be such that the differential impedance
is 85 Ω ± 10%.
• The PCIE_REXT contacts should be connected to a 200 Ω 1% resistor to ground. The trace length
between the pin and the resistor should be minimized. The resistor value is defined within the data
sheet and should determine the exact resistor value.
• Route traces over continuous planes (power and ground). Avoid split planes, plane slots, or
anti-etch.
• Maintain the parallelism (skew matched) between differential signals; these traces should be the
same overall length.
• Keep signals with traces as short as possible.
• Route signals with a minimum amount of corners. Use 45-degree turns instead of 90-degree turns.
• Do not create stubs or branches.
• Maintain symmetry of differential pair routing.
3.7.2 PCI Express coupling lane

Based on our development design, we have the following coupling signal schema. Consult the PCISig
documentation for detailed information.
• DC-coupled Rx signals with 0 Ω resistors
• AC-coupled Tx signals with 0.1µF capacitors
3.7.3 Additional resources for PCI Express signal routing

recommendations
For more information about, PCI Express signal routing recommendations, see the following.
• Freescale Hardware Design Considerations for PCI Express® and SGMII
(http://www.freescale.com/files/training_presentation/TP_HARDWARE_DESIGN_PCI_SMGIII
.pdf)
• PCISig, PCI Express Base Specification.
• PCISig, PCI Express Card Electromechanical Specification.
• PCISig, PCSIG Board Design Guidelines for PCI Express™ Architecture.
• PCI Express Basics: Developing Physical Design Rules for PCIe
(http://www.mentor.com/products/pcb-system-design/multimedia/pcie-basics-webinar)
3.8 HDMI recommendations

Use the following recommendations for the HDMI.
• HDMI differential pairs should have a impedance of 100 Ω in all paths to the connector.
• It is highly recommended to use an HDMI transmitter port protection for ESD, level shifting,
isolation, overcurrent and backdrive protecion.
3.9 SATA recommendations

Use the following recommendations for the SATA.
• SATA differential pairs should have a differential impedance of 100 Ω.
• Each differential pair should be length matched to ± 5 mils.
• Follow standard high-speed differential routing rules for signal integrity.
3.10 LVDS recommendations

Use the following recommendations for the LVDS.
• Follow standard high-speed differential routing rules for signal integrity.
• Each differential pair should be length matched to ± 5 mils.
• LVDS differential pairs should have a differential impedance of 100 Ω.
3.11 USB recommendations

Use the following recommendations for the USB.
• Route the high speed clocks and the DP and DM differential pair first.
• Route DP and DM signals on the top or bottom layer of the board
• The trace width and spacing of the DP and DM signals should be such that the differential
impedance is 90 Ω.
• Route traces over continuous planes (power and ground).
— They should not pass over any power/GND plane slots or anti-etch.
— When placing connectors, make sure the ground plane clearouts around each pin have ground
continuity between all pins.
• Maintain the parallelism (skew matched) between DP and DM; these traces should be the same
overall length.
• Do not route DP and DM traces under oscillators or parallel to clock traces and/or data buses.
• Minimize the lengths of high speed signals that run parallel to the DP and DM pair.
• Keep DP and DM traces as short as possible.
• Route DP and DM signals with a minimum amount of corners. Use 45-degree turns instead of
90-degree turns.
• Avoid layer changes (vias) on DP and DM signals. Do not create stubs or branches.
3.12 Impedance signal recommendations

Use the following table as a reference when you are updating or creating constraints in your software PCB
tool to set up the impedance and the correct trace width.
Table 3-8. Impedance signal recommendations
Layout
Signal Group Impedance
Tolerance (±)
All signals, unless specified 50 Ω SE 10%
PCIe Diff signals 85 Ω Diff 10%
USB Diff signals 90 Ω Diff 10%
Diff signals: 100 Ω Diff 10%

LVDS, SATA, HDMI, DDR, MIPI (CSI & DSI), MLB, Phy IC to Ethernet
Connector
The following figure shows the dimensions of a stripline and microstrip pair. Figure 3-32 shows the
differential pair routing.
Figure 3-31. Microstrip and stripline differential pair dimensions
Figure 3-32. Differential pair routing
• The space between two adjacent differential pairs should be greater than or equal to twice the space
between the two individual conductors.
• The skew between LVDS pairs should be within the minimum recommendation (± 100 mil).
3.13 Reference resistors

NOTE
The reference resistor and the connection should be placed away from noisy
regions. Noise induced on it may impact the internal circuit and degrade the
interface signals.
3.14 ESD and radiated emissions recommendations

The PCB design should use six or more layers, with solid power and ground planes. The recommendations
for ESD immunity and radiated emissions performance are as follows:
• All components with ground chassis shields (USB jack, buttons, and so forth) should connect the
shield to the PCB chassis ground ring.
• Ferrite beads should be placed on each signal line connecting to an external cable. These ferrite
beads must be placed as close to the PCB jack as possible.
NOTE
Ferrite beads should have a minimum impedance of 500 Ω at 100 MHz with
the exception of the ferrite on USB_5V.
• Ferrite beads should NOT be placed on the USB D+/D– signal lines as this can cause USB signal
integrity problems. For radiated emissions problems due to USB, a common mode choke may be
placed on the D+/D– signal lines. However, in most cases, it should not be required if the PCB
layout is satisfactory. Ideally, the common mode choke should be approved for high speed USB use
or tested thoroughly to verify there are no signal integrity issues created.
• It is highly recommended that ESD protection devices be used on ports connecting to external
connectors. See the reference schematic (available at www.freescale.com) for detailed information
about ESD protection implementation on the USB interfaces.
• If possible, stitch all around the board with vias with 100 mils spacing between them connected to
GND planes with exposed solder mask to improve EMI.
3.15 Component placement recommendations

Adhere to the following recommendations when placing components.
• Place components such that short and/or critical routes can be easily laid out.
— Critical routes determine component location.
— Orient devices to facilitate routes (minimize length and crossovers).
• Consider placing the following pairings adjacent.
— i.MX and DDR
— PHY and associated jack
— Jack and CODEC input
— Bluetooth® (or other RF) and antenna
3.16 Reducing skew and phase problems in deferential pairs traces

Differential pair technology has evolved to require more stringent checking in the area of phase control.
This is evident on the higher data rates associated with parallel buses such as PCI Gen 2, DDR, LVDS, or
Ethernet. In the simplest of terms, Diff Pair technology sends opposite and equal signals down a pair of
traces. Keeping these opposite signals in phase is essential to assuring that they function as intended.
Figure 3-33 and Figure 3-34 show two examples of static routing where a match is achieved without
needing to tune one element of the differential pair.
Figure 3-33. Yellow traces diff pairs 1
The following figure shows the addition of a delay trace to one element of the differential pair to avoid
length mismatch (which reduces skew and phase problems). The green box marks the detail.
Figure 3-34. Small bumps added to the shorter differential pair
Having this delay reduces skew and phase problems.
3.17 Guideline for power net electrical performance

The following figure shows the simulated open-circuit impedance of the SABRE AI platform power nets.
The board number is 700-27142, layout version is LAY-27142_C-1004.
The graph is provided as impedance guidance for the various power nets relative to ground. Freescale
recommends that a user’s board be at or below the impedances curves shown. For example, consider the
100 MHz point on the VDD_ARM_CAP curve at 0.065 ohms. User layouts at 100 MHz should have
impedance from VDD_ARM_CAP to ground of 0.065 ohms or less.
Figure 3-35. Power Net Open-Circuit Impedance – 10 kHz to 100 MHz
Figure 3-36. Power net open-circuit impedance-10 kHz to 10 GHz
Chapter 4
4.1 Power management requirements overview
This chapter provides the power requirements for the following i.MX 6 series families of processors:
• i.MX 6Quad
• i.MX 6Dual
• i.MX 6DualLite
• i.MX 6Solo
4.1.1 Voltage domains overview

These chips have several voltage domains that may need to be supplied with different voltages depending
on system needs. The chip internal regulators and its complementary PMIC PF0100 provide a complete
and simple way to supply each voltage domain with different voltages when needed. Section 4.4,
“Connection diagrams,” shows the internal regulators and the connections to PF0100.
4.1.2 PF0100 overview

PF0100 consists of the following components used to supply the i.MX6 voltage domains as well as other
blocks on the system:
• 4 buck regulators
• 1 boost regulator
• 8 LDOs
The default PF0100 power-up sequence is programmed to fit the requirements of the i.MX 6 series families
of processors. However, the PF0100 can be adjusted to meet the specific requirements for system
applications by using the one time programmable (OTP) feature.
4.2 Requirements for a generic interface between chip and PF0100

Table 4-1 shows the generic interface between the chip and PF0100, using a suitable power-up sequence.
For more info about PF0100 functionality and the i.MX 6 series families of processors’ power
requirements, see the product data sheets.
Table 4-1. Interface between the chip and PF0100
Supply
reg
Voltage Generated Power up
Voltage rail Supply reg current Notes
(V) by sequence
capability
(A)
VDDARM_IN SW1A/B 1.35 2.5 PF0100 1 Short these together with a shunt
for quad core operation. Cut shunt
VDDARM23_IN for dual core operation and connect
VDDARM23_IN to GND
VDDARM_CAP Note 1 (1.2) — i.MX — Short these together with a shunt

for quad core operation. Cut shunt
VDDARM23_CAP for dual core operation and connect
VDDARM23_IN to GND
VDDSOC_IN SW1C 1.325 2.0 PF0100 1 VDDARM_IN and VDDSOC_IN

supplies can be shorted together
and run off of one fewer switcher
for cost-sensitive platforms.
VDDSOC_CAP Note 1 (1.1) — i.MX Short these together.

VDDSOC_CAP is the output of an
i.MX6 internal LDO that can supply
more voltage domains as indicated
VDD_CACHE_CAP below in this table.
VDDPU_CAP Note 1 (1.2) — i.MX — —
VDDHIGH_IN SW2 3.0 2.0 PF0100 2 —

VDDHIGH_CAP Note 1 (2.5) — i.MX VDDHIGH_CAP is the output of an
i.MX6 internal LDO that can supply
more voltage domains as indicated
below in this table.
VDD_SNVS_IN VSNVS 3.0 400 μA PF0100 0 According to table 13 of the chip

data sheet, VDD_SNVS_IN can
draw up to 1 mA depending on the
application. For those cases, an
external regulator is needed
because the PF0100 VSNVS
regulator supplies 400 μA.
NVCC_RGMII SW3A/B 1.5 1.25 PF0100 3 —
NVCC_DRAM SW3 can be configured from 0.4 to

3.3 V so that the right voltage can
be chosen for the respective DDR
technology.
DRAM_VREF VREFDDR or 0.5×SW3 0.01 or 1.0 PF0100 3 —

SW4
Table 4-1. Interface between the chip and PF0100 (continued)
Supply
reg
(V) by sequence
capability
(A)
NVCC_CSI SW4, VGEN4 or 1.8–3.3 1.0 or 0.35 PF0100 or 5 Depending on system needs, these
external external voltage domains can be supplied
NVCC_EIM 0, 1, 21 regulator regulator together or independently with
NVCC_ENET equal or different voltages and
regulators. Be sure to account for
NVCC_GPIO the current needs of the domains
and the current capability of the
NVCC_LCD
regulator when making this
NVCC_NANDF decision.
NVCC_SD1, 2
NVCC_SD3
NVCC_JTAG
NVCC_PLL_OUT Note 2 1.1 — i.MX — —
NVCC_MIPI VDDHIGH_CAP 2.5 — i.MX — —
NVCC_LVDS2P5 VDDHIGH_CAP 2.5 — i.MX — This supply also powers the

pre-drivers of the DDR IO pins.
Therefore, it must always be
provided, even when LVDS is not
used.
USB_OTG_VBUS SWBST 5.0 0.6 PF0100 — • In Host configuration,

USB_OTG_VBUS can be fed
from the SWBST output of the
PF0100.
• In device configuration,
USB_OTG_VBUS is the
external host that provides this
voltage.
USB_H1_VBUS SWBST 5.0 0.6 PF0100 — Connect to VBUS pin of USB

connector
VDDUSB_CAP Note 2 (3.0) — i.MX — —
SATA_VP VDDSOC_CAP 1.1 — i.MX — —
SATA_VPH VDDHIGH_CAP 2.5 — i.MX — —
HDMI_VP VDDSOC_CAP 1.1 — i.MX — —
HDMI_VPH VDDHIGH_CAP 2.5 — i.MX — —
PCIE_VP VDDSOC_CAP 1.1 — i.MX — —
PCIE_VPTX — — —
Table 4-1. Interface between the chip and PF0100 (continued)
Supply
reg
(V) by sequence
capability
(A)
PCIE_VPH VDDHIGH_CAP 2.5 — i.MX — —
1 For Solo and DualLite chips, these 3 voltage rails should be connected together to the same voltage value.
Quad and Dual chips can support three different EIM power rails.
2 These voltage domains are supplied by i.MX6 internal regulators.
3 VGEN6, which comes up in sequence 4, does not power any i.MX rails and therefore is not shown. See Table 4-2 for VGEN6.
The following table shows the PF0100 regulators that are available to supply the rest of the system
circuitry.
Table 4-2. PF0100 regulators for other system circuitry
Supply Output voltage(V) Step size (mV) Maximum Load current (mA)
SW41 0.5 × SW3A_OUT, 0.4 – 3.3 25/50 1000
VGEN1 0.75 – 1.5 50 100
VGEN2 0.75 – 1.5 50 250
VGEN3 1.8 – 3.3 100 100
VGEN41 1.8 – 3.3 100 350
VGEN5 1.8 – 3.3 100 100
VGEN6 1.8 – 3.3 100 200
1 In Table 4-1, it was recommended to supply the NVCC_x voltage domains with SW4 or VGEN4. Depending on the decision,
one of them may not be available to supply the rest of the system circuitry.
4.3 i.MX6 internal regulators

These chips have been equipped with 7 internal regulators that simplify the power supply scheme of the
system. The following table shows the regulators’ power requirements. See Section 4.4, “Connection
diagrams,” for the distribution and connections of these LDOs.
Table 4-3. Internal regulator power requirements
LDO Output voltage (V) Output current (mA)
LDO_ARM 1.1 —
LDO_SOC 1.2 —
LDO_PU 1.1 —
Table 4-3. Internal regulator power requirements (continued)
LDO Output voltage (V) Output current (mA)
LDO_2P5 2.5 350
LDO_1P1 1.1 —
LDO_SNVS 1.1 —
LDO_USB 3.0 50
4.4 Connection diagrams
Figure 4-1. i.MX6x/PF0100 connection diagram, 1 of 2
1
Choose the pullup voltage for the I2C lines based on the I2C channel chosen. For example, for the I2C3 channel,
the corresponding voltage domain is NVCC_GPIO.
Figure 4-2. i.MX6x/PF0100 connection diagram, 2 of 2
4.5 Video power recommendations

VDD_PU_CAP is the supply for the internal video processing units (VPU). For video intensive operations,
the VPU requires a lot of power and may undergo large swings of instantaneous current requirements.
Therefore, the power supply to the VPU must be designed to handle relatively large surges of current at
high frequencies from the original source to the processor input for power (VDD_SOC_IN) and at the
output of the internal regulator for VPU operations (VDD_PU_CAP). The following list provides
recommendations for each specific point along the current supply path. It may be necessary to implement
all of these recommendations to ensure that one particular point along the supply path does not become a
current choke point.
• The voltage with which VDD_SOC_IN will be fed must have a maximum tolerance of ± 25 mV.
PF0100's SW1C is already designed with this tolerance. Care must be taken if the design uses a
different regulator.
• VDD_SOC_CAP and VDD_PU_CAP bulk capacitance must be equal to 22 μF so that start up
current through the on board LDOs is reduced.
• These bulk capacitors must be very close to the VDD_SOC_CAP and VDD_PU_CAP pins
respectively and the connecting traces must be as thick as the design allows so the ability of being
a bulk capacitor for high speed operations is not limited.
• VDD_SOC_IN requires 66 μF of bulk capacitance because it supplies power for both
VDD_SOC_CAP and VDD_PU_CAP.
Chapter 5
This chapter provides a reference table of the root clock default speed and a list of the i.MX modules
available to exit stop mode.
5.1 Root clocks

Clock connectivity is described in the “System Clocks Connectivity” section in the CCM chapter of the chip
reference manual. This section contains a series of tables that describe the clock inputs of each module and
which clock is connected to it.
NOTE
In some cases, a clock is associated with an external interface and is sourced
from a pad (mainly through IOMUX) and not from the CCM. Such clocks do
not appear in the clock connectivity table. They are found in the “External
Signals and Pin Multiplexing” chapter.
Clock gating is done with the low power clock gating (LPCG) module based on a combination of the clock
enable signals. For information about how the clock gating signals are logically combined, see the LPCG
section in the CCM chapter of the chip reference manual.
Table 5-1 lists the available clock sources and the default frequencies that are configured by design. In
some cases, users need to divide the clock inside the module when the maximum frequency is used in order
to meet the protocol requirements. CCM (the clock controller module) generates and drives the clock
sources.
For information about how the root clocks are generated, see the clock generation diagrams in the CCM
chapter of the chip reference manual.
Table 5-1. Clock roots
Clock Root Name (from CCM) Description Default frequency [MHz]
ARM_CLK_ROOT ARM core clock 792
MMDC_CH0_CLK_ROOT Multi Mode DDR Controller 528
AHB_CLK_ROOT AMBA Bus 132
IPG_CLK_ROOT Inter-packet Gap 66
PERCLK_CLK_ROOT 66
Table 5-1. Clock roots (continued)
Clock Root Name (from CCM) Description Default frequency [MHz]
USDHC1_CLK_ROOT Ultra Secured Digital Host 198

Controller
USDHC2_CLK_ROOT 198
USDHC3_CLK_ROOT 198
USDCH4_CLK_ROOT 198
SSI1_CLK_ROOT Synchronous Serial Interface 63.525
SSI2_CLK_ROOT 63.525
SSI3_CLK_ROOT 63.525
GPU2D_AXI_CLK_ROOT 2D Graphics Processing Unit 270
GPU3D_AXI_CLK_ROOT 270
PCIE_AXI_CLK_ROOT PCI Express 270
VDO_AXI_CLK_ROOT Video Data Order Adapter 270
AXI_CLK_ROOT Advanced eXtensible 270

Interface
IPU1_HSP_CLK_ROOT IPU High-Speed 264

Processing Clock
IPU2_HSP_CLK_ROOT 264
5.2 Waking the core up from stop mode

The following modules can wake the core up from stop mode.
• CAN
• ECSPI
• EIM
• ENET
• EPIT
• GPC
• GPIO
• GPT
• I2C
• KPP
• PCIE
• SDMA
• UART
• USB
Chapter 6
This chapter provides users with the basic information required to use the IOMUX system design aid
(IOMux.exe). The IOMUX design aid facilitates the assignment of internal signals to external device
balls/pins by helping users:
• Record signal assignments for the supported i.MX device
• Identify conflicts, allowing them to be resolved in real time
• Add notes or comments for each signal to the list of recorded assignments
• Generate C code to configure the IOMUXC registers according to the user’s design
• Move signals to different modules to order configuration code into logical functions
Users can save design configurations for future use and/or export them for use in schematics or software
source code as supplementary documentation of a system.
The following figure shows a screenshot of the IOMUX application window with various areas labeled.
Figure 6-1. IOMUX tool for the i.MX6 families of applications processors
6.1 Compatibility across the i.MX 6 series families of processors

The IOMUX registers for the i.MX 6 series families of processors are compatible as follows:
• The i.MX 6Quad family is compatible with the i.MX 6Dual family.
• i.MX 6DualLite family is compatible with the i.MX 6Solo family.
• i.MX 6Quad/6Dual families are not software compatible with i.MX 6DualLite/6Solo families.
Therefore, you cannot use code generated for i.MX 6Quad/6Dual for i.MX 6DualLite/6Solo and vice
versa.
6.2 Application requirements

The IOMUX application requires that the following be installed:
• Microsoft Windows XP or newer
• Microsoft’s .NET Framework, .NET Framework to 4.0 or newer.
6.3 IOMUX tool version

The IOMUX application i.MX 6Quad_6Dual IOMux Tool v3.2.1 supports the following devices in all
available package variations:
• i.MX 6Dual
• i.MX 6Quad
• i.MX 6Solo
• i.MX 6DualLite
6.4 IOMUX tool location

To obtain the IOMUX tool for the chip, consult your Freescale sales representative or download the
IOMUX tool from www.freescale.com.
Note that the IOMUX tool must be version v3.3 or later.
6.5 Learning to use the IOMUX tool

Consult the IOMUX user’s manual file inside the package for a detailed walkthrough of how to use the
IOMUX tool.
Chapter 7
This chapter explains how to configure JTAG tools for debugging. The JTAG module is a standard JEDEC
debug peripheral. It provides debug access to important hardware blocks, such as the ARM® processor
and the system bus, which can give users access and control over the entire chip.
To prevent JTAG manipulation while allowing access for manufacturing tests and software debugging, the
i.MX 6 series processor incorporates a secure JTAG controller for regulating JTAG access. The secure
JTAG controller provides four different JTAG security modes, which are selected by e-Fuse configuration.
For more information about the security modes, see the “Security” section in the “System JTAG Controller
(SJC)” chapter of the relevant i.MX6 chip reference manual.
7.1 JTAG tool requirements

To use JTAG tools, your system must have the following:
• Windows based PC
• RVDS v4.1 package or newer
• RealView ICE box connected to your computer
Freescale recommends making the JTAG port accessible during platform initial validation bring-up and
for software debugging. It is accessible in all development kits from Freescale.
Multiple tools are available for accessing the JTAG port for tests and software debugging. Freescale
recommends use of the ARM JTAG tools for compatibility with the ARM core. However, the JTAG chain
as described in the following sections should work with non-ARM JTAG tools. For more information
about configuring non-ARM tools, contact the third party tool vendor for support.
7.2 Extra JTAG functionality

Additional CoreSight debug components, such as trace machines using DS-5 debug software and
DSTREAMER hardware, can be used for extra JTAG functionality. However, they are not mandatory for
a basic configuration and are beyond the scope of this document.
NOTE
There is no option for using RVDS at its version at time of publication (4.1)
because it does not support PTM (i.MX 6 series trade module).
7.3 Updating your RealView ICE

Before using the RealView ICE for JTAG debugging, ensure you have the most up-to-date version
available. To update your RealView ICE, perform the following steps:
1. Launch the RVI Update utility by using the following path:
Start → Programs → ARM → RealView ICE v4.1 → RealView ICE Update
2. Connect to the ICE by selecting it from the list, as shown in the following figure.
Figure 7-1. Connecting to ICE
NOTE
The ICE must be disconnected from any other target at this step.
3. Select the firmware update from the upside menu: RVI → Install Firmware Update
4. Select the following file (or an equivalent more recent version):
C:\Program Files\ARM\RVI\Firmware\4.2\23\ARM-RVI-4.3.0-1-base.rvi
5. Select “Continue” from the install update window and wait until the update is complete.
Figure 7-2. Install update window
6. RVI automatically reboots.

7. Upon reconnecting to the RVI, you should see version number 4.3.0 build 1 or a later version
number. The exact version name should match with the version number installed in step 4 (see the
following figure).
Figure 7-3. RVI window after reconnecting
7.4 Defining the JTAG chain

To define the JTAG chain for an ARM Cortex®-A9 based chip, perform the following steps:
1. Find Freescale_iMX6 Q.rvs at the following location: …/My
Documents\ARM\rvconfig\platformFiles
NOTE
Be sure to use this path exactly, or the tool-chain configuration will not be
available from the Debugger-Connect to Target.
Contact your sales representative or go to www.freescale.com to obtain your

copy of the Freescale_iMX6 Q.rvs file.
2. Run RealView Debugger by using the following path: Start → Programs → ARM → RealView
Development Suite v4.1 → RealViewDebugger v4.1
3. Select Connect to Target in the RealView Debugger upside menu: Target → Connect To Target
4. Press Add near RealView ICE.
Figure 7-4. Adding your ICE
5. Select your ICE from the list and press Connect (see Figure 7-1).
6. In the new window, choose Select Platform…
7. Expand the “Freescale” list and select iMX6 Q.
8. Save the file (File → Save).
9. Close the window.
After finishing this procedure, you should see the following screen:
Figure 7-5. RealView debugger screenshot
Add the correct amount of Cortex A-9 cores desired to access your CPUs.
7.5 Reading a register with RealView Debugger v4.1

To read a register, perform the following steps:
1. Open the RealView Debugger 4.1 and connect to the target, as shown in the following figure.
Figure 7-6. Connecting to the target
2. You are now at rvdebug.brd; if you have successfully completed your setup, it looks like the
following screenshot:
Figure 7-7. Establishing a connection to the core
3. Establish the connection to the core of your choice by using the Connect icon or the shortcut
CTRL+N.
You now have a new RVI configuration with four Cortex-A9 targets and the RealView Debugger up and
running. You can now use the RealView Debugger window to access a register, as shown in the
following figure.
Figure 7-8. Accessing a register
Figure 7-8 shows an example of using the RealView Debugger to access the IOMUX register
IOMUXC_SW_MUX_CTL_PAD_GPIO_0, whose address is 0x020E0220 and whose default value after
reset is 0x5.
7.6 CoreSight Base address references

The CoreSight base addresses are as follows:
• For the i.MX 6Quad
— CPU#0: 0x82150000
— CPU#1: 0x82152000
— CPU#2: 0x82154000
— CPU#3: 0x82156000
• For the i.MX6 Solo, CPU#0: 0x82150000.
• For the i.MX 6Dual and 6DualLite
— CPU#0: 0x82150000
— CPU#1: 0x82152000
Chapter 8
This chapter provides recommendations for avoiding typical mistakes when bringing up a board for the
first time. These recommendations consist of basic techniques that have proven useful in the past for
detecting board issues and addressing the three most typical bring-up pitfalls: power, clocks, and reset. A
sample bring-up checklist is provided at the end of the chapter.
8.1 Using a current monitor to avoid power pitfalls

Excessive current can cause damage to the board. Avoid this problem by using a current-limiting
laboratory supply set to the expected typical main current draw (at most). Monitor the main supply current
with an ammeter when powering up the board for the first time. You can use the supply's internal ammeter
if it has one. By monitoring the main supply current and controlling the current limit, any excessive current
can usually be detected before permanent damage occurs.
8.2 Using a voltage report to avoid power pitfalls

Using incorrect voltage rails is a common power pitfall. To help avoid this mistake, create a basic table
called a voltage report prior to bringing up your board. This table helps validate that all the supplies are
reaching the expected levels.
To create a voltage report, list the following:
• Your board voltage sources
• Default power-up values for the board voltage sources
• Best location on the board to measure the voltage level of each supply
Carefully determine the best measurement location for each power supply to avoid a large voltage drop
(IR drop) on the board, which causes inaccurate current values to be measured. The following guidelines
help produce the best current measurements:
• Measure closest to the load (in this case the i.MX6 processor).
• Make two measurements: the first after initial board power-up and the second while running a
heavy use-case that stresses the i.MX6 processor.
Ensure that the supplies that are powering the i.MX6 meet the DC electrical specifications as listed in your
chip-specific data sheet.
The following table illustrates how a sample voltage report table helps detect errors. The shaded cells in
the PMIC LDO2 row call your attention to the difference in the expected value and measured value, which
indicates a potential problem with that power rail.
Table 8-1. Sample voltage report
Expected Measured Measured

Source Net name Comment
(V) (V) point
Main 5V0 5.0 5.103 C5.1 —
3.3 V discrete reg 3V3_DELAYED 3.35 3.334 SH1 Requires LDO3 to enable
PMIC Switcher 1 VDDARM 1.375 1.377 SH2 —
PMIC Switcher 2 VDDSOC 1.375 1.376 SH3 —
PMIC Switcher 3 1V5_DDR 1.5 1.501 SH4 —
PMIC LDO1 1V8 1.8 1.802 TP9 —
PMIC LDO2 2V5 2.5 0.3 TP5 —
VREFDDR 0V75_REFDDR 0.75 0.751 C8.1 50% of 1V5_DDR
Coin Cell 3V0_STBY 3.0 3.006 TP1 —
i.MX6 VDDARM_CAP 1.1 1.114 C6.1 —
i.MX6 VDDHIGH_CAP 2.5 2.515 SH5 —
i.MX6 VDDSNVS_CAP 1.0 1.016 TP2 —
8.3 Checking for clock pitfalls

Problems with the external clocks are another common source of board bring-up issues. Ensure that all of
your clock sources are running as expected.
The XTALI/XTALO and the RTC_XTALI/RTC_XTALO clocks are the main clock sources for 24 MHz
and 32 kHz reference clocks respectively on the i.MX6. Although not required, the use of low jitter
external oscillators to feed CLK1_P/N or CLK2_P/N on the i.MX6 can be an advantage if low jitter or
special frequency clock sources are required by modules driven by CLK1_P/N or CLK2_P/N. See the
CCM chapter in your i.MX6 chip reference manual for details. If a 32.768 kHz crystal is not connected to
the i.MX6, an on-chip ring oscillator is automatically used for the low-frequency clock source.
When checking crystal frequencies, use an active probe to avoid excessive loading. A parasitic probe
typically inhibits the 32.768 kHz and 24 MHz oscillators from starting up. Use the following guidelines:
• RTC_XTALI clock is running at 32.768 kHz (can be generated internally or applied externally).
• XTALI/XTALO is running at 24 MHz (used for the PLL reference).
• CLK1_P/N/CLK2_P/N can be used as oscillator inputs for low jitter special frequency sources.
• CLK1_P/N and CLK2_P/N are optional.
In addition to probing the external input clocks, you can check internal clocks by outputting them at the
debug signals CLKO1 and CLKO2 (iomuxed signals). See the CCM chapter in the chip reference manual
for more details about which clock sources can be output to those debug signals. JTAG tools (see
Chapter 7, “Configuring JTAG Tools”) can be used to configure the necessary registers to do this.
8.4 Avoiding reset pitfalls

Follow these guidelines to ensure that you are booting using the correct boot mode.
• During initial power on while asserting the POR_B reset signal, ensure that 24 MHz clock is active
before releasing POR_B.
• Follow the recommended power-up sequence specified in the i.MX6 data sheet.
• Ensure the POR_B signal remains asserted (low) until all voltage rails associated with bootup are
on.
The GPIOs and internal fuses control how the i.MX6 boots. For a more detailed description about the
different boot modes, see the system boot chapter of the chip reference manual.
The following figures show two examples of the power-up sequence.
5V_MAIN
Feeds 3.3 V Reg
SNVS
SW1A/B
(VDDARM_IN)
POR_B
RESETBMCU
Figure 8-1. Power-up sequence example 1
SW1A/B
(VDDARM_IN)
SW1C
(VDDSOC_IN)
SW2
(VDDHIGH_IN)
VGEN2
(System 1.5V)
Figure 8-2. Power-up sequence example 2
8.5 Sample board bring-up checklist

Table 8-2 provides a sample board bring-up checklist. Note that the checklist incorporates the
recommendations described in the previous sections. Blank cells should be filled in during bring-up as
appropriate.
Table 8-2. Board bring-up checklist
Findings &
Checklist Item Details Owner
status
Note: The following items must be completed serially.
1. Perform a visual inspection. Check major components to make sure nothing has been
misplaced or rotated before applying power.
2. Verify all i.MX6 voltage rails. Confirm that the voltages match the data sheet’s
requirements. Be sure to check voltages not only at the
voltage source, but also as close to the i.MX6 as possible
(like on a bypass capacitor). This reveals any IR drops on
the board that will cause issues later.
Ideally all of the i.MX6 voltage rails should be checked, but
VDD_ARM_IN and VDD_SOC_IN are particularly important
voltages. These are the core logic voltages and must fall
within the parameters provided in the i.MX6 data sheet.
VDD_SNVS_IN, NVCC_JTAG, and NVCC_DRAM are also
critical to the i.MX6 boot up.
Note: NVCC_LVDS2V5 must be powered when using the
chip DDR interface. This power input is used as the
Pre-Driver power source for the DDR I/O pads.
Table 8-2. Board bring-up checklist (continued)
Findings &
Checklist Item Details Owner
status
3. Verify power-up sequence. Verify that power on reset (POR_B) is de-asserted (high)
after all power rails have come up and are stable. See the
i.MX6 data sheet for details about power-up sequencing.
4. Measure/probe input clocks (32 kHz, Without a properly running clock, the i.MX6 will not function
24MHz, others). properly.
5. Check JTAG connectivity (RV-ICE). This is one of the most fundamental and basic access points
to the i.MX6 to allow the debug and execution of low level
code.
Note: The following items may be worked on in parallel with other bring up tasks.
Access internal RAM. Verify basic operation of the i.MX6 in system. The on-chip
internal RAM starts at address 0090_0000h and is
256 Kbytes in density. Perform a basic test by performing a
write-read-verify to the internal RAM. No software
initialization is necessary to access internal RAM.
Verify CLKO outputs (measure and This ensures that the corresponding clock is working and
verify default clock frequencies for that the PLLs are working.
desired clock output options) if the board Note that this step requires chip initialization, for example
design supports probing of the CLKO via the JTAG debugger, to properly set up the IOMUX to
pin. output CLKO and to set up the clock control module to
Note: output the desired clock. See the reference manual for more
details.
Measure boot mode frequencies. Set This verifies the specified signals’ connectivity between the
the boot mode switch for each boot i.MX6 and boot device and that the boot mode signals are
mode and measure the following properly set.
(depending on system availability): See the “System Boot” chapter in the reference manual for
• NAND (probe CE to verify boot, details about configuring the various boot modes.
measure RE frequency)
• SPI-NOR (probe slave select and
measure clock frequency)
• MMC/SD (measure clock frequency)
Run basic DDR initialization and test 1. Assuming the use of a JTAG debugger, run the DDR
memory. initialization and open a debugger memory window pointing
to the DDR memory map starting address.
2. Try writing a few words and verify if they can be read
correctly.
3. If not, recheck the DDR initialization sequence and
whether the DDR has been correctly soldered onto the
board.
It is also recommended that users recheck the schematic to
ensure that the DDR memory has been connected to the
i.MX6 correctly.
Chapter 9
This chapter explains how to use the IBIS (input output buffer information specification) model, which is
an Electronic Industries Alliance standard for the electronic behavioral specifications of integrated circuit
input/output analog characteristics. The model is generated in ASCII text format and consists of multiple
tables that capture current vs. voltage (IV) and voltage vs. time (VT) characteristics of each buffer. IBIS
models are generally used to perform PCB-board-level signal integrity (SI) simulations and timing
analyses.
The IBIS model’s features are as follows:
• Supports fast chip-package-board simulation, with SPICE-level accuracy and faster than any
transistor-level model
• Provides the following for portable model data
— I/O buffers, series elements, terminators
— Package RLC parasitics
— Electrical board description
9.1 IBIS structure and content

An IBIS file contains the data required to model a component’s input, output, and I/O buffers behaviorally
in ASCII format. The basic IBIS file contains the following data:
• Header information regarding the model file
• Information about the component, the package’s electrical characteristics, and the pin-to-buffer
model mapping (in other words, which pins are connected to which buffer models)
• The data required to model each unique input, output, and I/O buffer design on the component
IBIS models are component-centric, meaning they allow users to model an entire component rather than
only a particular buffer. Therefore, in addition to the electrical characteristics of a component’s buffers, an
IBIS file includes the component’s pin-to-buffer mapping and the electrical parameters of the component’s
package.
9.2 Header Information

The first section of an IBIS file provides the basic information about the file and its data. The following
table explains the header information notation. Example 9-1 shows what header information looks like in
an IBIS file.
Table 9-1. Header Information
Keyword Required Description
[IBIS Ver] Yes Version of IBIS Specification this file uses.
[Comment char] No Change the comment character. Defaults to the pipe (|) character
[File Name] Yes Name of this file. All file names must be lower case. The file name extension for an IBIS file is .ibs
[File Rev] Yes The revision level of this file. The specification contains guidelines for assigning revision levels.
[Date] No Date this file was created
[Source] No The source of the data in this file. Data is taken from a simulation and validated on the board.
[Notes] No Component or file-specific notes.
[Disclaimer] No May be legally required
[Copyright] No The file’s copyright notice
Example 9-1. Header Information

[IBIS Ver] 4.2
[Comment Char] |_char
[File Name] 21x21_imx6q_autmtv_003.ibs
[File Rev] 003
[Date] Wed Mar 14 14:22:34 2012
[Source] FSL Viper 2012.03.14
[Notes]
9.3 Component and pin information

The second section of an IBIS file is where the data book information regarding the component’s pinout,
pin-to-buffer mapping, and the package and pin electrical parameters is placed. The following table
explains the component and pin information notation, and Example 9-2 shows what it looks like in an IBIS
file.
Table 9-2. Component and Pin Information
Keyword Required Comment
[Component] Yes The name of the component being modeled. Standard practice has been to use the industry
standard part designation. Note that IBIS files may contain multiple [Component] descriptions.
[Manufacturer] Yes The name of the component manufacturer
Table 9-2. Component and Pin Information (continued)
[Package] Yes This keyword contains the range (minimum, typical and maximum values) over which the
packages’ lead resistance, inductance, and capacitance vary (the R_pkg, L_pkg, and C_pkg
parameters).
[Pin] Yes This keyword contains the pin-to-buffer mapping information. In addition, the model creator can
use this keyword to list the package information: R, L, and C data for each individual pin (R_pin,
L_pin, and C_pin parameters).
[Package No If the component model includes an external package model (or uses the [Define Package Model]
Model] keyword within the IBIS file itself), this keyword indicates the name of that package model.
[Pin Mapping] No This keyword is used if the model creator wishes to include information on buffer power and ground
connections. This information may be used for simulations involving multiple outputs switching.
[Diff Pin] No This keyword is used to associate buffers that should be driven in a complementary fashion as a
differential pair.
[Model This keyword provides a simple means by which several buffers can be made optionally available
Selector] for simulation at the same physical pin of the component.
Example 9-2. Component and pin information

[Component] iMX6Q
[Manufacturer] FREESCALE
[Package]
| variable typ min max
R_pkg 0.3508997 0.0028730 0.895806
L_pkg 2.62395nH 0.07145nH 5.71558nH
C_pkg 3.89344pF 0.59645pF 29.3042pF
|
[Pin] signal_name model_name R_pin L_pin C_pin
A2 PCIE_REXT gpio 0.646088 4.19621nH 1.57274pF
A3 PCIE_TXM gpio 0.615102 4.09171nH 1.58877pF
...
[Pin Mapping] pulldown_ref pullup_ref

A2 GND PCIE_VPH
A3 GND PCIE_VPH
...
[Diff Pin] inv_pin vdiff tdelay_typ tdelay_min tdelay_max
A6 B6 NA NA NA NA
A10 B10 NA NA NA NA
A12 B12 NA NA NA NA
...
[Model Selector] ddr

ddr3_sel11_ds111_mio DDR, 1.5V, ddr3 mode, 34 Ohm driver impedance
rgmii_sel11_ds111_mio DDR, 2.5V, 31 Ohm driver impedance
9.4 Model information

The [Model] keyword starts the description of the data for a particular buffer. Table 9-3 shows the main
sets of parameters and keywords, composing the model definition.
Table 9-3. Model information
Keyword Comment
[Model Spec] General set of parameters for the model simulation.
[Receiver Thresholds] Threshold information for the different simulation cases.
[Temperature Range] The temperature range over which the min, typ and max IV and switching data have been
gathered.
[Voltage Range] The range over which Vcc is varied to obtain the min, typ and max pullup and power clamp data.
[Pulldown] IV information. For more details, see Section 9.4.1, “ IV information.”

[Pullup]
[GND_clamp]
[POWER_clamp]
[Ramp] VT information. For more details, see Section 9.4.2, “VT information.”
[Rising Waveform]
[Falling Waveform]
[Test Data] VT golden model information. For more details, see Section 9.4.3, “Golden Model VT information.”
[Rising Waveform Near]
[Rising Waveform Far]
[Falling Waveform Near]
[Falling Waveform Far]
[Test Load]
9.4.1 IV information
IV information is composed of four Current-over-Voltage tables: [Pullup], [Pulldown], [GND_clamp], and
[Power_clamp]. Each look-up table describes a different part of the IO cell model, as shown in Table 9-1.
Figure 9-1. Model IV parameters’ structure
9.4.2 VT information
The following table defines the keywords that provide the information about an output or I/O buffer, and
Example 9-3 shows what they look like in an IBIS file.
Table 9-4. Ramp and waveform keywords
[Ramp] Yes Basic ramp rate information, given as a dV/dt_r for rising edges and dV/dt_f for falling
edges, see the following equation.
dV 20 % to 80% voltage swing
------- = ----------------------------------------------------------------------------------
dt time taken to swing above voltage
Note: The dV value is the 20% to 80% voltage swing of the buffer when driving into
the specified load, R_load (for [Ramp], this load defaults to 50). For CMOS
drivers or I/O buffers, this load is assumed to be connected to the voltages
defined by the [Voltage Range] keyword for falling edges and to ground for
rising edges.
[Rising Waveform] No The actual rising (low to high transition) waveform, provided as a VT table.
[Falling Waveform] No The actual falling (high to low transition) waveform, provided as a VT table.
Example 9-3. Ramp and waveform keywords example

[Ramp]
| variable typ min max
dV/dt_r 0.4627/0.3456n 0.4326/0.4568n 0.4962/0.3030n
dV/dt_f 0.4546/0.3481n 0.4272/0.3918n 0.4774/0.3569n
R_load = 0.2400k
|
[Rising Waveform]
R_fixture= 0.2400k
V_fixture= 0.0
V_fixture_min= 0.0
V_fixture_max= 0.0
|time V(typ) V(min) V(max)
|
|0.0S 0.3369uV 12.4052uV 41.7335nV
|19.7866fS 0.6730uV 12.7375uV 0.3823uV
|20.8863fS 0.6917uV 12.7519uV 0.4013uV
|21.9489fS 0.7058uV 12.7657uV 0.4196uV
...
|
[Falling Waveform]
R_fixture= 0.2400k
V_fixture= 0.0
V_fixture_min= 0.0
V_fixture_max= 0.0
|time V(typ) V(min) V(max)
|
|0.0S 0.7711V 0.7211V 0.8270V
|0.3334nS 0.7711V 0.7211V 0.8270V
|0.3445nS 0.7711V 0.7211V 0.8269V
...
The [Ramp] keyword is always required, even if the [Rising Waveform] and [Falling Waveform] keywords
are used. However, the VT tables under [Rising Waveform] and [Falling Waveform] are generally
preferred to [Ramp] for the following reasons:
• VT data may be provided under a variety of loads and termination voltages
• VT tables may be used to describe transition data for devices as they turn on and turn off.
• [Ramp] effectively averages the transitions of the device, without providing any details on the
shapes of the transitions themselves. All detail of the transition ledges would be lost.
The VT data should be included under two [Rising Waveform] and two [Falling Waveform] sections, each
containing data tables for a Vcc-connected load and a Ground-connected load (although other loading
combinations are permitted).
The most appropriate load is a resistive value corresponding to the impedance of the system transmission
lines the buffer will drive (own impedance). For example, a buffer intended for use in a 60 Ω system is
best modeled using a 60 Ω load (R_fixture).
The following figure shows how to interpret the model data.
Figure 9-2. Model data interpretation
9.4.3 Golden Model VT information

Golden waveforms are a set of waveforms simulated using known ideal test loads. They are useful for
verifying the accuracy of behavioral simulation results against the transistor level circuit model from
which the IBIS model parameters originated.
The following figure shows a generic test load network.
| V_term1
| o-----------o
| | |
| \ \ receiver_model_name
| ______ / / ______
| | | NEAR Rp1_near \ \ Rp1_far FAR | |
| | |\ | / / | |\ |
| | | \ | Rs_near Ls_near | _____ | Ls_far Rs_far | | \ |
| | | >-|---o--/\/\--@@@@--o----o--O_____)--o----o--@@@@--/\/\--o---|-| > |
| | | / | | | | Td | | | | | / |
| | |/ | | C1_near | \ Zo \ | C2_far | | |/ |
| |______| === === / / === === |______|
| | C2_near | \ \ | C1_far |
| | | / / | |
| | | | V_term2 | | |
| o--------------o o-----------o o--------------o
| | Rp2_near Rp2_far |
| GND GND
Figure 9-3. Generic test load network
The following table explains the golden waveform keywords.

Table 9-5. Golden waveform keywords
[Test Data] No • Provides a set of golden waveforms and references the conditions under which
they were derived.
• Useful for verifying the accuracy of behavioral simulation results against the
transistor level circuit model from which the IBIS model parameters originated.
[Rising Waveform Near] Yes Current-Over-Voltage tables, for far and near portions of the golden model as
[Rising Waveform Far] described by Figure 9-3.
[Falling Waveform Near]
[Falling Waveform Far]
[Test Load] Yes • Defines a test load network and its associated electrical parameters for reference
by golden waveforms under the [Test Data] keyword.
• If Test_load_type is Differential, the test load is a pair of the circuits shown in . If the
R_diff_near or R_diff_far subparameter is used, a resistor is connected between
the near or far nodes of the two circuits.
• If Test_load_type is Single_ended, R_diff_near and R_diff_far are ignored.
9.5 Freescale naming conventions for model names and usage in

i.MX6 IBIS file
The model names are defined per each [Model selector]. The models may differ from each other by having
different parameters—such as voltage, drive strength, mode of operation, and slew rate. The mode of
operation, drive strength, and slew rate parameters are programmable by software.
9.5.1 [Model Selector] ddr

The “ddr” model type supports both the DDR and the RGMII protocol signals.
9.5.1.1 DDR [Model Selector]

“ddr” models exist for DDR3, DDR3L, DDR3U and LPDDR2 protocols.
This model has the following parameters:
• DDR protocol
• DDR IO type
• Drive strength
• ODT enable/disable
The IBIS model name is composed from the parameters’ values in two ways, as follows:
• Without active ODT circuit:
<ddr protocol>_sel<ddr_type>_ds<drive_strength>_mio
• With active ODT circuit:
<ddr protocol>odt_t<ODT_value>_sel<ddr_type>_mi
DDR write models ("_mio" suffix) have no simulated ODT, as ODT is disabled during write. Write
models' DS parameter is meaningful and changes to describe the different levels of drive strength.
DDR read models ("_mi" suffix) have no meaningful DS parameter, as no driving happens during read.
Read models’ ODT parameter is meaningful and changes to describe different levels of ODT impedance.
DDR Protocol Selected according to the used DDR. DDR IO voltage level is selected
accordingly.
DDR IO Type Controlled by the IOMUXC_SW_PAD_CTL_GRP_DDR_TYPE[19:18] register
in IOMUXC (IOMUX controller) DDR_SEL bits, to select between DDR3 &
LPDDR2.
Drive strength Controlled by bits [5:3] (DSE) of the following registers in IOMUXC (IOMUX
controller):
IOMUXC_SW_PAD_CTL_PAD_DRAM_SDCLK_x (2 registers)
IOMUXC_SW_PAD_CTL_PAD_DRAM_CAS
IOMUXC_SW_PAD_CTL_PAD_DRAM_RAS
IOMUXC_SW_PAD_CTL_PAD_GRP_ADDDS
IOMUXC_SW_PAD_CTL_PAD_DRAM_RESET
IOMUXC_SW_PAD_CTL_PAD_DRAM_SDCKEx (2 registers)
IOMUXC_SW_PAD_CTL_PAD_DRAM_SDODTx (2 registers)
IOMUXC_SW_PAD_CTL_PAD_GRP_CTLDS
IOMUXC_SW_PAD_CTL_PAD_DRAM_SDQSx (8 registers)
IOMUXC_SW_PAD_CTL_PAD_DRAM_BxDS (8 registers)
IOMUXC_SW_PAD_CTL_PAD_DRAM_DQMx (8 registers)
ODT value Controlled by bits [18:16], [14:12], [10:8], and [6:4] in MPODTCTRL register of
MMDC.
Example 9-4. [Model Selector] DDR in IBIS file
ddr3_sel11_ds111_mio DDR, 1.5V, ddr3 mode, 34 Ohm driver impedance

...
lpddr2_sel10_ds111_mio LPDDR, 1.2V, lpddr2 mode, 34 Ohm driver impedance
lpddr2_sel10_ds110_mio LPDDR, 1.2V, lpddr2 mode, 40 Ohm driver impedance
...
See the register description in the IOMUXC chapter in the chip reference manual for further details about
this model.
9.5.1.2 RGMII
• RGMII voltage
• Drive strength
The IBIS model name is composed from the parameters’ values as follows:
rgmii_sel11_ds<drive_strength>_mio
Voltage Level IBIS currently supports only the 2.5 V option. 2.5 V is applied to NVCC_RGMII.
No further register programing is required.
Drive strength Controlled by bits [5:3] (DSE) of the following registers in IOMUXC (IOMUX
controller):
IOMUXC_SW_PAD_CTL_PAD_RGMII_TXC
IOMUXC_SW_PAD_CTL_PAD_RGMII_TX_CTL
IOMUXC_SW_PAD_CTL_PAD_RGMII_TDx (4 registers)
IOMUXC_SW_PAD_CTL_PAD_RGMII_RXC
IOMUXC_SW_PAD_CTL_PAD_RGMII_RX_CTL
IOMUXC_SW_PAD_CTL_PAD_RGMII_RDx (4 registers)
IO Type Regardless of the voltage level, he ddr_sel of
IOMUXC_SW_PAD_CTL_GRP_DDR_TYPE_RGMII should always be set to
‘11’.
Example 9-5. [Model Selector] RGMII in IBIS file

...
9.5.2 [Model Selector] gpio

• Voltage level
• Drive strength
• Slew rate
• Speed
The IBIS model name is composed from parameters’ values as follows:
gpio<voltage_level>_ds<drive_strength>_sr<slew_rate(1 bit)><speed(2 bits)>_mio
Voltage level For i.MX6 chips, there are no user configurations for the voltage level because the
GPIO cell senses the NVCC and auto-configures itself accordingly. The IBIS user
can choose between high and low voltage by selecting a different model at [Model
Selector].
Drive strength Controlled by the DSE bits (bits [5:3]) in the
IOMUXC_SW_PAD_CTL_PAD_<pad name>.
Slew rate Controlled by the SRE bit (bit 0) in the IOMUXC_SW_PAD_CTL_PAD_<pad
name>.
Speed Controlled by the SPEED bits (bits [7:6]) in the

IOMUXC_SW_PAD_CTL_PAD_<pad name>.
Example 9-6. [Model Selector] gpio in IBIS file
[Model Selector] gpio
gpiohv_ds111_sr111_mio GPIO, 3.3V,extra drive, fast slew, max frequency
gpiohv_ds111_sr110_mio GPIO, 3.3V,extra drive, fast slew, fast frequency
gpiohv_ds111_sr101_mio GPIO, 3.3V,extra drive, fast slew, medium frequency
gpiohv_ds111_sr100_mio GPIO, 3.3V,extra drive, fast slew, slow frequency
...
gpiolv_ds111_sr111_mio GPIO, 1.8V,extra drive, fast slew, max frequency
gpiolv_ds111_sr110_mio GPIO, 1.8V,extra drive, fast slew, fast frequency
gpiolv_ds111_sr101_mio GPIO, 1.8V,extra drive, fast slew, medium frequency
...
See the register description in the IOMUXC chapter in the chip reference manual for further details about
this model.
9.5.3 [Model Selector] lvds

A single model is available for LVDS, as no configurable parameters exist for this IO model.
The LVDS model is available not only for the LVDS port signals, but also for the general purpose CLK1_x
and CLK2_x, who share the same IO model.
Example 9-7. [Model Selector] lvds in IBIS file
[Model Selector] lvds
lvds_mio LVDS, Vos = 1.2V, Voh = 1.375, Vol = 1.025, Vovdd = 2.5
...
9.5.4 [Model Selector] mlb

The following two models are available for MLB, as no configurable parameters exist for this IO model.
Example 9-8. [Model Selector] lvds in IBIS file
[Model Selector] mlb
mlb_sig_data_mio MLB, Signal/Data Transceiver, Vod = 400, Vid = 100
mlb_clk_mi MLB, CLK Receiver, Vid = 100
...
9.5.5 [Model Selector] USB

At the time of publication, i.MX6 IBIS rev 3 does not contain the USB model. It is expected to be
published in a future revision.
9.5.6 List of pins not modeled in the i.MX6 IBIS file

The following table provides a list of analog or special interface pins that are not modeled in the i.MX6
IBIS file.
Table 9-6. i.MX6 pins not supported by IBIS
ANALOG MIPI HDMI PCIe SATA Analog USB
RTC_XTALI CSI_CLK0M HDMI_CLKM PCIe_RXM SATA_RXM USB_H1_DN
RTC_XTALO CSI_CLK0P HDMI_CLKP PCIe_RXP SATA_RXP USB_H1_DP
XTALI CSI_D0M HDMI_D0M PCIe_TXM SATA_TXM USB_H1_VBUS
XTALO CSI_D0P HDMI_D0P PCIe_TXP SATA_TXP USB_OTG_DN
ZQPAD CSI_D1M HDMI_D1M USB_OTG_DP
CSI_D1P HDMI_D1P USB_OTG_VBUS
CSI_D2M HDMI_D2M USB_OTG_CHD_B
CSI_D2P HDMI_D2P
CSI_D3M HDMI_DDCCEC
CSI_D3P HDMI_HPD
DSI_CLK0M
DSI_CLK0P
DSI_D0M
DSI_D0P
DSI_D1M
DSI_D1P
NOTE
In rev3 of the i.MX6 IBIS, some of the above unsupported pins are
described as “GPIO” cells. These are no more than placeholders and cannot
be used for signal modeling.
9.6 Quality assurance for the IBIS models

The IBIS models are validated against the IBIS specification, which provides a way to objectively measure
the correlation of model simulation results with reference transistor-level spice simulation or
measurements.
Correlation The process of making a quantitative comparison between two sets of I/O buffer
characterization data, such as lab measurement vs. structural simulation or
behavioral simulation vs. structural simulation.
Correlation Level A means for categorizing I/O buffer characterization data based on how much the
modeling engineer knows about the processing conditions of a sample component
and which correlation metric he or she used.
All models (GPIO, DDR, LVDS, MLB) have passed the following checks:
• IBISCHK without errors or unexplained warnings
• Data for basic simulation checked
• Data for timing analysis checked
• Data for power analysis checked
• Correlated against Spice simulations
Validation reports can be provided upon demand.
9.7 IBIS usage

Freescale board designers used the i.MX6Q IBIS model with the Hyperlynx tool by Mentor Graphics. The
HyperLynx version used was HyperLynx v8.1.1 + Update 2.
Effective board design results achieved after loading:
• i.MX6Q IBIS model.
• Companion IC IBIS models.
• Board model in HyperLynx format.
Board simulations for various GPIO, LVDS, and DDR signals were then run.
9.8 References
Consult the following references for more information about the IBIS model.
• IBIS Open Forum (http://www.eda.org/ibis/)
The IBIS Open Forum consists of EDA vendors, computer manufacturers, semiconductor vendors,
universities, and end-users. It proposes updates and reviews, revises standards, and organizes
summits. It promotes IBIS models and provides useful documentation and tools.
• IBIS specification (http://eda.org/pub/ibis/ver5.0/)
Chapter 10
10.1 Overview
The i.MX manufacturing tool is designed to program firmware onto storage devices such as NAND or eSD
through the EVK and preload the data area with media files in an efficient and convenient manner. It is
intended for Freescale Semiconductor customers or their OEMs who plan to mass manufacture
i.MX-based products.
The application is not designed to test the devices or to diagnose manufacturing problems. Devices
initialized with this application still need to be functionally verified.
10.2 Feature summary

The tool includes the following features:
• Continuous operation—operations automatically begin with the connection of a new device, and
multiple operations such as update and copy can be linked together seamlessly.
• Enumeration—static-ID firmware loaded into RAM in recovery-mode prevents Windows® from
enumerating every device.
• AutoPlay—various Windows® ‘pop-up’ application and status messages, such as Explorer in
Windows® XP and Windows 7.
In addition, the following characteristics improve the tool’s ease of use:
• An independent process bar is set up for each physical USB port.
• The tool begins processing with the connection of the first device detected and allows users to
replace each device after completion instead of needing to wait for all devices to complete.
• The tool uses color-based indicators to indicate the work status on each of the ports.
— Blue indicates the device is being processed.
— Green indicates the device was successfully processed and that the programmed device can be
replaced with a new one independent of the of the device’s progress.
— Red indicates the device failed to process.
10.3 Other references

For more detailed information about the manufacturing tool, see the following documents included in the
manufacturing tool release package. Contact your local Freescale sales office for assistance obtaining
documents if needed:
• For detailed information about how to use the manufacturing tool, see Manufacturing Tool V2
Quick Start Guide.
• For detailed information about how to script the processing operations of the manufacturing tool,
see the Manufacturing Tool V2 UCL User Manual.
• For information about how to generate the manufacturing tool firmware for Linux and Android,
see Manufacturing Tool V2 Linux or Android Firmware Development Guide.
• For the change list and known issues, see Manufacturing Tool V2 Release Notes.
Chapter 11
11.1 BSDL overview
Boundary scan description language (BSDL) is used for board-level testing after components have been
assembled. The interface for this test uses the JTAG pins. The definition is contained within IEEE Std
1149.1.
11.2 How BSDL functions

A BSDL file defines the internal scan chain, which is the serial linkage of the IO cells, within a particular
device. The scan chain looks like a large shift register, which provides a means to read the logic level
applied to a pin or to output a logic state on that pin. Using JTAG commands, a test tool uses the BSDL
file to control the scan chain so that device-board connectivity can be tested.
For example, when using an external ROM test interface, the test tool would do the following:
1. Output a specific set of addresses and controls to pins connected to the ROM
2. Perform a read command and scan out the values of the ROM data pins.
3. Compare the values read with the known golden values.
Based on this procedure, the tool can determine whether the interface between the two parts is connected
properly and does not contain shorts or opens.
11.3 Downloading the BSDL file

The BSDL file for each i.MX processor is stored on the Freescale website upon product release. Contact
your local sales office or fields applications engineer to check the availability of information prior to
product releases.
11.4 Pin coverage of BSDL

Each pin is defined as a port within the BSDL file. You can open the file with a text editor (like Wordpad)
to review how each pin will function. The BSDL file defines these functions as shown:
-- PORT DESCRIPTION TERMS
-- in = input only
-- out = three-state output (0, Z, 1)
-- buffer = two-state output (0, 1)
-- inout = bidirectional
-- linkage = OTHER (vdd, vss, analog)
The appearance of ”linkage” in a pin’s file implies that the pin cannot be used with boundary scan. These
are usually power pins or analog pins that cannot be defined with a digital logic state.
11.5 Boundary scan operation

The boundary scan operation is controlled by:
• BOOT_MODE0, BOOT_MODE1, and JTAG_MOD pins
• On-chip Fuse bits
The JTAG_MOD pin state controls the selection of JTAG to the core logic or boundary scan operation.
See the following references for further information:
• The “System JTAG Controller (SJC)” chapter in the chip reference manual for the definitions of
the JTAG interface operations.
• The “JTAG Security Modes” section in the same chapter for an explanation of the operation of the
e-Fuse bit definitions in the following table.
• The “Fusemap” chapter in the chip reference manual the fusemap tables.
Table 11-1. System considerations for BSDL
Pin name Logic state Description
JTAG_MOD 1 IEEE 1149.1 JTAG compliant mode
BOOT_MODE[1:0] [0:0] Boot From Fuses

[0:1] Serial Downloader
[1:0] Internal Boot (Development)
POR_B 1 Power On Reset for the device
e‐Fuse bits
JTAG_SMODE[1:0] [0:0] JTAG enable mode
[0:1] Secure JTAG mode
SJC_DISABLE 0 Secure JTAG Controller is enabled
11.6 I/O pin power considerations

The boundary scan operation uses each of the available device pins to drive or read values within a given
system. Therefore, the power supply pin for each specific module must be powered in order for the IO
buffers to operate. This is straightforward for the digital pins within the system.
NOTE
BSDL was only tested at 1.8 V.
SATA and PCIe are not digital interfaces, but these modules provide built-in support for the IEEE 1149.6
extension for AC testing of their pins. Therefore, these modules must also be powered when utilizing a
scan chain that contains the pins from these modules, or the scan chain does not function properly.
Chapter 12
12.1 Overview
This chapter provides supporting instructions for the use of the i.MX 6 series Ethernet RMII interface.
NOTE
This chapter only covers the required hardware and register settings.
Modifications to the Ethernet driver or its initialization code are beyond its
scope. For this information, see your BSP documentation.
12.2 Configuring the RMII signal connections
Figure 12-1. Reference schematic, part 1 of 2
Figure 12-2. Reference schematic, part 2 of 2
Figure 12-1 provides a reference schematic, which shows the connections required to use the RMII
interface. These signal connections are generally self-explanatory or explained in the chip reference
manual. However, there are some required modifications.
12.3 Generating the reference clock

The Ethernet MAC needs to have a reference clock, which can be generated in one of the following three
ways:
• On chip clock generator
• By an external oscillator
• By the RMII PHY
Note that the pin labeled “ENET_REF_CLK” in Figure 12-2 is only required by the full MII interface. It
is not used by the RMII interface.
12.4 Generating the reference clock on chip

There are two possible pins that can either source or sink the reference clock: GPIO_16 and
RGMII_TX_CTL. The GPIO_16 pin is the preferred choice because it has the advantage of being in a
high voltage IO domain, which means it can be used at the standard 3.3 V IO voltage levels.
RGMII_TX_CTL should only be used if pin function loadings are such that GPIO_16 is unavailable.
12.4.1 Using the GPIO_16 pin to generate the reference clock

The following figure shows how to configure the GPIO_16 pad to generate the reference clock.

GPR1[21] (ENET_CLK_SEL) = 1
IOMUX
GPIO_16
ANATOP
PHY
X

ENET

IOMUXC_SW_MUX_CTL_PAD_GPIO16[4] SION =

Figure 12-3. Using the GPIO_16 pad
Note that the block labeled “ANATOP” is really the Analog Ethernet PLL. See your chip reference manual
for its control register figure. Bits 1–0 of CCM_ANALOG_PLL_ENETn control the frequency fed to the
Ethernet MAC and should be set to 01b to obtain 50 MHZ.
To use GPIO_16 as the RMII reference clock source, use the following:
• Set mode to ALT2 (MUX_MODE[2:0] = 010).
• Set the SION bit. Note that this is not required because the function setting controls the signal path,
but it is good practice as it reminds the user that the clock needs to fed back into the Ethernet MAC.
• For the internal clock case, set GPR1[21].
GPR1[21] controls the actual clock source.
• When cleared, it obtains the ENET Tx reference clock from a pad (external OSC for both external
PHY and internal controller).
• When set, it obtains the ENET Tx reference clock from ANATOP (loopback through pad) and
sends out to the external PHY.
The Daisy Chain register also needs to be set correctly to force the input to use the right pin. Note that there
is a subtle difference between the i.MX 6Quad/6Dual families and the i.MX 6DualLite/6Solo families that
affects the correct setting:
• For i.MX6Quad/6Dual—To use GPIO_16, set DAISY[0] = 1. If DAISY[0] is left at its reset value
(0b), RGMII_TX_CTL will be used instead. See the Select Input Register
(IOMUX_ENET_REF_CLK_SELECT_INPUT) in the IOMUXC chapter of the i.MX
6Dual/6Quad Reference Manual (IMX6DQRM), available at www.freescale.com
• For i.MX 6Solo/6DualLite—To use GPIO_16, leave the value of DAISY[0] at is reset value (0b).
To use RGMII_TX_CTL, set DAISY[0] = 1.
For further information, see the “DAISY (IOMUXC_ENET_REF_CLK_SELECT_INPUT)” section in
the “IOMUX Controller (IOMUXC)” chapter of your chip reference manual.
Note that while you can use the default pad settings as shown in the “SW_PAD_CTL
(IOMUXC_SW_PAD_CTL_PAD_GPIO16)” section in the IOMUX controller chapter, it may be
desirable to set the Slew Rate Field, SRE[0], to Fast (SRE[0] = 1).
12.4.2 Using RGMII_TX_CTL to generate the reference clock

RGMII_TX_CTL is in the RGMII IO voltage domain, which has a maximum voltage of 1.9 V (2.5 V in
silicon revision 1.2). Therefore, to use RGMII_TX_CTL, you must use a level shifter to match the RMII
PHY voltage levels. This is why GPIO_16 is preferred.
To use RGMII_TX_CTL, set the following:
• In the Daisy Chain register, ensure DAISY[0] is cleared; note that this is its default setting after
reset.
• Set RGMII_TX_CTL pad mux register to ALT7.
• If desired, set the SION bit as discussed in Section 12.4, “Generating the reference clock on chip”.
• The RGMII_TX_CTL pad control register, IOMUXC_SW_PAD_CTL_PAD_RGMII_TX_CTL,
does not have a slew rate control bit. Slew rate can be controlled by judicious choice of output
drive strength in the DSE field, bits 5:3
12.5 Using an external clock

GPR1[21] (ENET_CLK_SEL) = 0
IOMUX
GPIO_16
ANATOP
PHY
X

Oscillator
ENET

IOMUXC_SW_MUX_CTL_PAD_GPIO16[4] SION = 1

Figure 12-4. External clock configuration (external oscillator shown)
Figure 12-4 shows how to use an external clock. This configuration is almost identical when using an
external oscillator or the RMII PHY to supply a clock. The only required modification to an RMII PHY
instead of the external oscillator is to clear GPR1[21] (GPR1[21] = 0) to select the external clock input.
All other settings remain the same.
Appendix A
This appendix provides a complete list of the development platforms that are available from Freescale to
support the following i.MX 6 families of processors:
• i.MX 6Quad
• i.MX 6Dual
• i.MX 6DualLite
• i.MX 6Solo
You can use these tables as a quick guide for finding the best development platform for your needs. Note
that although these development platforms are based on a specific product family, they will work with any
of the i.MX product families listed above.
Table A-1. SABRE Board for Smart Devices
Version i.MX used i.MX 6Quad
Schematic PN and Rev. 170-27392
• 1 Gbyte DDR3
• SPI Nor
• eMMC Socket
• SD Card Socket
• SATA
• LVDS Ports 0 & 1
• HDMI
• Port of CSI CMOS Sensor (camera)
• MIPI CMOS Sensor
• MIPI Display Port
Features • Parallel Display Port
• TouchScreen
• Audio CODEC
• Ethernet
• 3 Axis Accelerometer
• Barometer
• Digital eCompass
• Aux SDIO Socket
• CAN Port (optional)
• Mini PCIe
•
Available at www.freescale.com/imxsabre on Freescale website.

Quick Start Guide
Freescale Semiconductor A-1
Table A-1. SABRE Board for Smart Devices (continued)

Schematic

Layout
Table A-2. SABRE Platform for Smart Devices
• i.MX 6Quad
Version i.MX used
• i.MX 6DualLite
Schematic PN and Rev. 170-27392
• 1Gbyte DDR3
• SPI Nor
• eMMC Socket
• SD Card Socket
• SATA
• LVDS Ports 0 & 1
• HDMI
• Port of CSI CMOS Sensor (camera)
• MIPI CMOS Sensor
• MIPI Display Port
• EPD Display Port
• Parallel Display Port
Features
• TouchScreen
• Audio CODEC
• Ethernet
• Light Sensor
• 3 Axis Accelerometer
• Barometer
• Digital eCompass
• Aux SDIO Socket
• CAN Port (optional)
• Mini PCIe
• GPS Receiver
• Battery Charger Options
Quick Start Guide Available at freescale.com/imxsabre on the Freescale website.
Schematic Available at freescale.com/imxsabre on the Freescale website.
Layout Available at freescale.com/imxsabre on the Freescale website.
A-2 Freescale Semiconductor
Appendix B
Revision History
Table B provides a revision history for this document.
Table B-1. Document Revision History
Rev.
Date Substantive Change(s)
Number
1 06/2013 • Added preface to book; Chapter 1, “About This Book”

• Chapter 2, “Design Checklist”:
- Table 2-1, “DDR recommendations,” recommendation 2, first bullet: Changed “0.1 F” to “0.1 μF.”
- Table 2-1, “DDR recommendations,” recommendation 3: Regarding LPDDR2, changed from
“DRAM_RESET pull is not necessary” to “DRAM_RESET should be left unconnected.”
- Standardized nomenclature for pin names and signal names. For further details on recent
standardization of signal names for the i.MX 6 series, see the i.MX 6 Series Standardized Signal
Name Map (EB792).
• Section 11.6, “I/O pin power considerations”: Added note specifying that BSDL was only tested at
1.8 V.
• Updated and added content to Table 2-6, Table 2-7 and Table 2-9
• Added new row to table Table 2-10
• Updated third column of all rows in Table 2-13
• Updated row 5 of table Table 2-14
• Corrected references throughout the book
• Added footnote to figure Figure 3-5.
• Added footnote to Table 4-1
• Updates to Figure 12-3. and Figure 12-4.
0 10/2012 Initial release.
Freescale Semiconductor BB-1
Revision History
B-2 Freescale Semiconductor

IMX6 DDR3 Rules PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

IMX6 DDR3 Rules PDF

Uploaded by

Copyright:

Available Formats

Hardware Development Guide

for i.MX 6Quad, 6Dual, 6DualLite,

Web Support: information in this document.

Document Number: IMX6DQ6SDLHDG

1.1 Overview.......................................................................................................................... 1-1

2.1 Design checklist overview ............................................................................................... 2-1

3.1 Basic design recommendations........................................................................................ 3-1

4.1 Power management requirements overview .................................................................... 4-1

5.1 Root clocks ...................................................................................................................... 5-1

6.1 Compatibility across the i.MX 6 series families of processors........................................ 6-2

7.1 JTAG tool requirements ................................................................................................... 7-1

8.1 Using a current monitor to avoid power pitfalls .............................................................. 8-1

9.1 IBIS structure and content ............................................................................................... 9-1

10.1 Overview........................................................................................................................ 10-1

11.1 BSDL overview ............................................................................................................. 11-1

12.1 Overview........................................................................................................................ 12-1

1.2 Devices supported

1.3 Essential reference

1.4 Suggested reading

1.4.1 General Information

• Computer Organization and Design: The Hardware/Software Interface (Second Edition), by

1.5 Related documentation

1.7 Signal conventions

1.8 Acronyms and abbreviations

Table 1: Definitions and acronyms

ARM® Advanced RISC machines processor architecture

BGA Ball grid array package

BOM Bill of materials

BSDL Boundary scan description language

CAN Flexible Controller Area Network peripheral

CCM Clock Controller Module

Table 1: Definitions and acronyms (continued)

DDC VESA Data Display Channel

DDR Dual data rate DRAM

DDR3 DDR3 DRAM

DDR3L Low voltage DDR3 DRAM

DDR3U Ultra low voltage DDR3 DRAM

DRAM Dynamic random access memory

DSI MIPI display serial interface

ECSPI Enhanced Configurable SPI peripheral

EDID Extended Display Identification Data

EIM External Interface Module

ENET 10/100/1000-Mbps Ethernet MAC peripheral

EPIT Enhanced Periodic Interrupt Timer peripheral

ESR Equivalent series resistance (of a crystal)

FSL Freescale Semiconductor

GPC General Power Controller

GPIO General-purpose input/output

HDCP High-bandwidth Digital Content Protection

HDMI High-definition multimedia interface

I2C Inter-integrated circuit interface

IBIS Input output buffer information specification

IOMUX i.MX6 chip-level I/O multiplexing

JTAG Joint Test Action Group

LDB LVDS Display bridge

LDO Low drop-out regulator

LPCG Low power clock gating

LPDDR2 Low-power DDR2 DRAM

LVDS Low-voltage differential signaling

MLB MediaLB 150 peripheral

MMDC Multi Mode DDR Controller

ODT On-die termination

OTP One-time programmable